Akustička Svojstva I Filogenetski Odnosi Glavoča Porodice Gobiidae (Teleostei; Gobiiformes)

Home , Benthophiloides, Caffrogobius, Caspian goby, Chlamydogobius, Chromogobius, Corcyrogobius

Akustička svojstva i filogenetski odnosi glavoča porodice Gobiidae (Teleostei; Gobiiformes)

Horvatić, Sven

Doctoral thesis / Disertacija

2021

Degree Grantor / Ustanova koja je dodijelila akademski / stručni stupanj: University of Zagreb, Faculty of Science / Sveučilište u Zagrebu, Prirodoslovno-matematički fakultet

Permanent link / Trajna poveznica: https://urn.nsk.hr/urn:nbn:hr:217:019844

Rights / Prava: In copyright

Download date / Datum preuzimanja: 2021-10-05

Repository / Repozitorij:

Repository of Faculty of Science - University of Zagreb PRIRODOSLOVNO-MATEMATIČKI FAKULTET

BIOLOŠKI ODSJEK

Sven Horvatić

AKUSTIČKA SVOJSTVA I FILOGENETSKI ODNOSI GLAVOČA PORODICE GOBIIDAE (TELEOSTEI; GOBIIFORMES)

DOKTORSKI RAD

Zagreb, 2021

PRIRODOSLOVNO-MATEMATIČKI FAKULTET

BIOLOŠKI ODSJEK

Sven Horvatić

AKUSTIČKA SVOJSTVA I FILOGENETSKI ODNOSI GLAVOČA PORODICE GOBIIDAE (TELEOSTEI; GOBIIFORMES)

DOKTORSKI RAD

Mentor:

Prof. dr. sc. Davor Zanella

Zagreb, 2021

FACULTY OF SCIENCE

DEPARTMENT OF BIOLOGY

Sven Horvatić

ACOUSTIC PROPERTIES AND PHYLOGENETIC RELATIONSHIPS OF GOBIES FROM THE FAMILY GOBIIDAE (TELEOSTEI; GOBIIFORMES)

DOCTORAL DISSERTATION

Supervisor:

Prof. Davor Zanella

Zagreb, 2021

Ovaj doktorski rad izrađen je u Laboratoriju za kralješnjake Zoologijskog zavoda Biološkog odsjeka Prirodoslovno-matematičkog fakulteta Sveučilišta u Zagrebu pod vodstvom prof. dr. sc. Davora Zanelle u sklopu Sveučilišnog poslijediplomskog doktorskog studija Biologije na Biološkom odsjeku Prirodoslovno-matematičkog fakulteta Sveučilišta u Zagrebu. Eksperimentalni dio ovog doktorskog rada obavljen je na Sveučilištu u Zagrebu, Prirodoslovno-matematičkom fakultetu, Biološkom odsjeku (Hrvatska), dok je dio bioakustičkih i anatomskih istraživanja obavljen na Sveučilištu Ca' Foscari u Veneciji (Italija) i Sveučilištu u Liègeu (Belgija).

Informacije o mentoru

ZANELLA, Davor (Zagreb, 19. travnja 1973.), redoviti profesor, 235863

U Zagrebu je završio osnovnu i srednju školu. Maturirao je 1992. g. na V. gimnaziji, a iste je godine upisao studij biologije, smjer profesor biologije na PMF-u u Zagrebu. Diplomirao je 1997. g. Od 2000. g. zaposlen je u Zoologijskom zavodu kao znanstveni novak. Godine 2003. obranio je magistarski rad, a 2007. i doktorsku disertaciju. Na mjesto docenta izabran je 2009., 2013. u izvanrednog profesora, a 2019. godine u redovitog profesora. Usavršavao se iz ekologije slatkih voda (određivanje kvalitete vode pomoću biotičkog indeksa) i populacija riba riječnih tokova i jezera u Laboratoriju za ihtiologiju i plankton Sveučilišta u Trstu, boravio je u Tromsu, u Norveškoj na stručnom usavršavanju iz područja ihtiologije u sklopu projekta "Management of Freshwater Fisheries of Bordering Rivers". Do 2009. g. vodi praktikume na preddiplomskom studiju iz kolegija Kralježnjaci, kao i terensku nastavu iz zoologije. Od 2009. g. nositelj je ili sunositelj kolegija Vertebrata, Upravljanje prirodnim populacijama, Ekologija i zaštita prirode, Zoologija 3 (Kralježnjaci), Ihtiologija, Akvakultura i ribarstvo slatkih voda i Akvakultura za studente preddiplomskog i diplomskog studija različitih smjerova. Sudjeluje u terenskoj nastavi iz navedenih kolegija. Voditelj je jednog doktorskog rada, 15 diplomskih radova, dva magistarska i 19 završnih radova. Njegova znanstvena djelatnost obuhvaća istraživanja raznolikosti i staništa slatkovodnih riba Hrvatske, posebno endemskih vrsta (inventarizaciju i utvrđivanje stupnja ugroženosti te bioloških i ekoloških značajki pojedinih vrsta). Drugo je područje njegova znanstvenog rada usmjereno na proučavanje prirodnih populacija te njihovom upravljanju. Objavio je kao autor ili koautor 58 znanstvenih radova (47 u časopisima indeksiranim u CC, 11 u časopisima indeksiranim u drugim sekundarnim i tercijarnim bazama). Koautor je tri knjige te jednog poglavlja u knjizi. Sudjelovao je na međunarodnim skupovima s 46 priopćenja te domaćim skupovima s 12 priopćenja. Sudjelovao je na brojnim međunarodnim i domaćim skupovima sa 60-ak priopćenja. Aktivno je sudjelovao u izvedbi triju znanstvenih projekata. Bio je i suradnik na dva međunarodna projekta.

Zahvale

Iskreno zahvaljujem mentoru prof. dr. sc. Davoru Zanelli na pruženoj pomoći tijekom doktorskog studija. Njegova mi je podrška, uz dobronamjerne komentare i prijedloge, mnogo pomogla. Također, zahvaljujem neposrednom mentoru sa Sveučilišta Ca' Foscari iz Venecije, prof. dr. sc. Stefanu Malavasiju što me tijekom studija uveo u neistražen i intrigantan svijet bioakustike. Njegova su mi nazočnost i znanstvena kompetencija bili od neizmjerne važnosti.

Posebno zahvaljujem povjerenstvu za obranu doktorskih radova, izv. prof. dr. sc. Perici Mustafiću, prof. dr. sc. Zlatku Liberu i izv. prof. dr. sc. Marcelu Kovačiću, na stručnim komentarima i dobronamjernim prijedlozima povezanima sa sadržajem ovog doktorskog rada.

Veliko hvala kolegama iz laboratorija za biologiju kralješnjaka i zaštitu prirode – Ivani, Luciji Raguž, Luciji Ivić, Zoranu, Marku, Perici, Romanu, Siniši, Miloradu – na pomoći i udijeljenim savjetima. Također, hvala vam što ste mi pokazali kako znanost, uz sve striktne norme, može biti i zabavna. Posebno zahvaljujem kolegi Zoranu, koji je tijekom studija, a pogotovo za vrijeme izrade ove disertacije, pokazao iznimno strpljenje i otvorenost za pomoć. Hvala svim koautorima znanstvenih radova ove disertacije na stručnoj pomoći i srdačnoj spremnosti na suradnju. Posebno zahvaljujem Lindi na iznimnoj pomoći s lekturom radova, kao i Maji Marčić na stručnoj lekturi ove disertacije..

Iako je teško nabrojiti sve, neizmjerno zahvaljujem prijateljima iz skupine „Tko nas tjera“ „4πCassa“ i „Vijeće mudraca“, Soži, Nataši, Milošu, Niki, Sonji i Loparanima. Hvala vam što ste mi, kada god je trebalo, udijelili iskren i kvalitetan životni savjet, ali i rasteretili um mnogim pošalicama i dogodovštinama.

Neizmjerno zahvaljujem svojoj najbližoj obitelji, majci Branki, ocu Braci i bratu Tinu. Puno vam hvala na pruženoj i neupitnoj podršci tijekom studiranja i života općenito. Hvala vam na svakom dobronamjernom savjetu, kritici ili komentaru koji su me vratili na pravi put kada god je trebalo. Zahvaljujem bakama Jasminki i Miki, djedu Ivi i uji, ujni te Benjaminu i Bruni na svakom razgovoru i savjetu. Drago mi je da ste sudjelovali u svemu! I za kraj, zahvaljujem svojoj Ivani što mi je unijela osmijeh i neizrecivu sreću u život te unaprijedila rad svojim komentarima.

Sveučilište u Zagrebu Doktorski rad

Prirodoslovno-matematički fakultet

Biološki odsjek

AKUSTIČKA SVOJSTVA I FILOGENETSKI ODNOSI GLAVOČA PORODICE GOBIIDAE (TELEOSTEI; GOBIIFORMES)

SVEN HORVATIĆ

Prirodoslovno-matematički fakultet, Biološki odsjek

Glavoči su najbrojnijih skupina riba zrakoperki. Također, oni su jedna od akustički najistraživanijih skupina zrakoperki današnjice, sa 23 poznate vrste koje se glasaju. U sklopu ove disertacije, u četiri znanstvene publikacije prvi je put istražena, upotrebom bioakustičkih metoda i analiza, sposobnost stvaranja zvukova i vokalni repertoar glavoča linije Gobius Neogobius melanostomus, Neogobius fluviatilis i Ponticola kessleri, kao i Perccottus glenii (Odontobutidae). Nadalje, ovom su disertacijom istražene anatomske strukture uključene u proces produkcije zvukova kod rotana, a rezultati su pokazali kako se mehanizam najvjerojatnije temelji na kontrakcijama kranijalno-pektoralnih mišića. Upotrebom molekularnih analiza koje uključuju metode filogenetske rekonstrukcije s mitohondrijskim i jezgrenim genetičkim biljezima, rekonstruirani su odnosi vokalnih glavoča linije Gobius te istražena korelacija između genetičke i akustičke divergencije. Rezultati ove disertacije naglašavaju kako su zvukovi glavoča specifični za pojedinu vrstu, dok njihova interspecijska akustička divergencija prati genetičku, što pokazuje kako evolucija zvukova i genotipova prati isti obrazac.

(79 stranica, 8 slika, 207 literaturnih navoda, jezik izvornika: hrvatski)

Ključne riječi: glavoči, bioakustika, linija Gobius, filogenija, zvukovi

Mentor: prof. dr. sc. Davor Zanella

Ocjenjivači: izv. prof. dr. sc. Perica Mustafić

prof. dr. sc. Zlatko Liber

izv. prof. dr. sc. Marcelo Kovačić

University of Zagreb Doctoral thesis

Faculty of Science

Department of Biology

ACOUSTIC PROPERTIES AND PHYLOGENETIC RELATIONSHIPS OF GOBIES FROM THE FAMILY GOBIIDAE (TELEOSTEI; GOBIIFORMES)

SVEN HORVATIĆ

Faculty of Science, Department of Biology

Gobies are the most numerous fish group. They are also one of the most investigated acoustic cluster of actinopterygian fish nowadays, with 23 species vocally tested. In this dissertation, represented through four scientific publications, the acoustic communication and vocal repertoire was investigated in Gobius-lineage gobies Neogobius melanostomus, Neogobius fluviatilis and Ponticola kessleri, as well as in Perccottus glenii (Odontobutidae), using standardized bioacoustic methods and analyses. The presented dissertation investigated the anatomical structures involved in the sound production, and the conducted analyses indicate that the sonic mechanism could be based on the contractions of cranio-pectoral muscles. By using phylogenetic analyses with a multilocus approach, we reconstructed the interspecific relationships between vocal gobies from Gobius-lineage and correlated the acoustic and genetic divergences. Results of this dissertation emphasize that goby sounds are truly species-specific, while their interspecific acoustic divergence follows the genotypic divergence, which could indicate that acoustic signals in gobies from Gobius-lineage carry phylogenetic signal.

(79 pages, 8 figures, 207 references, original in Croatian)

Keywords: gobies, bioacoustics, Gobius-lineage, phylogeny, sounds

Supervisor: Dr. Davor Zanella

Reviewers: Dr. Perica Mustafić

Dr. Zlatko Liber

Dr. Marcelo Kovačić

Tablica sažetka

POPIS PUBLIKACIJA ...... I PROŠIRENI SAŽETAK ...... II THESIS SUMMARY ...... VI 1. UVOD ...... X 1.1. Glavoči, najbrojnija skupina riba zrakoperki ...... 1 1.2. Pregled najvažnijih filogenetskih istraživanja europskih glavoča linije Gobius (Gobiiformes, Gobiidae) ...... 7 1.3. Bioakustika, znanost o zvukovima u prirodi ...... 15 1.4. Bioakustika riba ...... 21 1.5. Pregled najvažnije literature o bioakustici glavoča ...... 26 2. ZNANSTVENE PUBLIKACIJE ...... XI 3. DISKUSIJA ...... XII 3.1. Glavoči linije Gobius – modelni organizmi u bioakustičkim istraživanjima ...... 38 3.2. Uvid u mehanizam za produkciju zvukova kod glavoča ...... 42 3.3. Akustička divergencija glavoča linije Gobius – povezanost s genetskom diferencijacijom? 46 4. ZAKLJUČAK ...... XIII 5. LITERATURA ...... 54 6. ŽIVOTOPIS ...... 67

POPIS PUBLIKACIJA

1. Horvatić S, Cavraro F, Zanella D i Malavasi S (2016) Sound production in the Ponto- Caspian goby Neogobius fluviatilis and acoustic affinities within the Gobius lineage: implications for phylogeny. Biological Journal of the Linnean Society 117: 564-573. 2. Horvatić S, Bem L, Malavasi S, Marčić Z, Buj I, Mustafić P, Ćaleta M, Zanella D (2019) Comparative analysis of sound production between the bighead goby Ponticola kessleri and the round goby Neogobius melanostomus: Implications for phylogeny and systematics. Environmental Biology of Fishes 102: 727-739. 3. Horvatić S, Malavasi S, Parmentier E, Marčić Z, Buj I, Mustafić P, Ćaleta M, Smederevac-Lalić M, Skorić S, Zanella D (2019) Acoustic communication during reproduction in the basal gobioid Amur sleeper and the putative sound production mechanism. Journal of Zoology 309: 269-279. 4. Horvatić S, Malavasi S, Vukić J, Šanda R, Marčić Z, Ćaleta M, Lorenzoni M, Mustafić P, Buj I, Raguž L, Ivić L, Cavraro F, Zanella D (2021) Acoustic divergence is correlated with phylogenetic distance in European gobies (Gobiidae; Gobius lineage): evolutionary insights. Predano na recenziju u Plos One.

PROŠIRENI SAŽETAK

Glavoči (porodice Gobiidae i Gobionellidae, prema Thacker, 2009) su najbrojnija skupina riba zrakoperki, s oko 1600 opisanih vrsta te mnoštvom neotkrivenih i kriptičkih skupina rasprostranjenih u gotovo svim akvatičkim ekosustavima (Nelson, 2006; Fricke i sur., 2020). Oni naseljavaju najraznolikija vodena staništa, kao što su duboka i hladna mora, potoci i rijeke te prijelazne bočate lagune i estuarije, a u plitkom infralitoralu dominiraju po brojnosti i raznolikosti vrsta. Zbog svojih bioloških karakteristika koje uključuju pridneni ili kriptobentički način života, speleofilni tip reprodukcije i teritorijalnost s izraženom roditeljskom brigom za potomke, glavoči su iznimno zanimljivi modelni organizmi u laboratorijskim i in situ bioakustičkim eksperimentima. Upravo su zato jedna od akustički najistraživanijih skupina riba današnjice (uz Pomacentridae, Batrachoidiidae, Gadidiae, Serrasalmidae), s 23 dosad poznate vrste koje se glasaju (oko 1,2% ukupne brojnosti prema Fricke i sur., 2020) ispitane u bioakustičkim eksperimentima (Zeyl i sur., 2016). Do danas se većina znanstvene pozornosti u bioakustičkim ispitivanjima pridavala glavočima linije Pomatoschistus (rodovi Knipowitschia, Economidichthys, Gobiusculus i Pomatoschistus) i atlantsko-mediteranskim skupinama linije Gobius (Zosterisessor, Padogobius i Gobius), dok je znanje o akustičkim signalima, vokalnom repertoaru i etološkom kontekstu produkcije zvukova bilo znatno neistraženo kod pontokaspijskih vrsta i glavoča u širem smislu (Gobioidei). Multidisciplinarna istraživanja navode kako vokalni glavoči iz dviju različitih porodica, Gobius paganellus (Gobiidae) i Pomatoschistus pictus (Gobionellidae), imaju (uz male osteološke razlike) karakterističan plan građe i jednake anatomske strukture u glavenom dijelu i oplećju, zbog čega njihovi osnovni elementi za produkciju zvukova najvjerojatnije slijede sličan evolucijski obrazac koji potencijalno dijele s ostalim sestrinskim skupinama kao što su Odontobutidae, Butidae i dr. (Parmentier i sur., 2013, 2017). Produkcija zvukova, izuzev navedene dvije porodice, istražena je još samo kod jedne vrste iz porodice Odontobutidae (Odontobutis obscura), a bioakustičko istraživanje potvrdilo je kako mužjaci te vrste stvaraju pulsatilni tip zvuka pomoću ždrijelnih zuba (Takemura, 1983). Stoga je detaljna bioakustička i anatomska analiza trebala istražiti vokalni repertoar i anatomske strukture uključene u proces stvaranja zvukova kod bazalnih pripadnika glavoča u širem smislu, kao što je primjerice rotan Perccottus glenii. Nadalje, filogenetski odnosi glavoča i dalje su prilično neistraženi iako novija filogenetska istraživanja navode kako se europski glavoči dijele, uz porodice, i na nekoliko evolucijskih skupina ili linija (Gobius i Aphia unutar potporodice Gobiine; Pomatoschistus unutar potporodice Gobionelline, prema Agorreta i sur.,

II 2013). Molekularne analize, tj. metode filogenetske rekonstrukcije temeljene na pristupu multilokus (više različitih mitohondrijskih i jezgrenih gena), naglašavaju kako unutar linije Gobius atlantsko-mediteranski rodovi Gobius, Padogobius i Zosterisessor dijele bliske sestrinske odnose s pontokaspijskom skupinom (rodovi Neogobius, Ponticola, Proterorhinus i dr.) rasprostranjenom u slijevovima Crnog, Azovskog i Aralskog mora te Kaspijskog jezera (Thacker i Roje, 2011; Agorreta i sur., 2013). Iako su odnosi ove skupine detaljnije istraženi koristeći isključivo molekularne ili morfološke metode, akustička komparativna analiza dosad nije provedena. Usto, niti jedno istraživanje dosad nije empirijski i kvantitativno povezalo akustičku divergenciju s genetičkom diversifikacijom kod vokalnih riba kako bi se istražilo slijede li one isti obrazac evolucije, tj. nose li zvukovi filogenetski signal zaslužan za razlikovanje vokalnih vrsta iz linije Gobius.

Cilj je ove disertacije bio proširiti znanje o vokalnom repertoaru (s akustičkim parametrima), anatomskim strukturama uključenima u proces stvaranja zvukova te filogenetskim odnosima vokalnih glavoča linije Gobius (Gobiiformes, Gobiidae) i rotana (Gobiiformes, Odontobutidae) koristeći tradicionalne bioakustičke metode i analize te metode filogenetske rekonstrukcije s više različitih genetičkih biljega (mtDNA i nDNA). U četiri znanstvene publikacije, cilj je disertacije bio istražiti i testirati ove znanstvene hipoteze: i. Filogenetska interspecijska analiza pontokaspijskih i atlantsko-mediteranskih glavoča iz linije Gobius, rekonstruirana na temelju kvalitativnih i kvantitativnih akustičkih parametara, upućuje na određen stupanj međusobne srodnosti između vokalnih vrsta, ii. Kod glavoča linije Gobius glasaju se samo mužjaci tijekom intraspecijskih ili interspecijskih interakcija, iii. Mehanizam za produkciju zvukova temelji se kod rotana Perccottus glenii (Odontobutidae) na kranio-pektoralnim mišićima, a pulsatilni zvukovi predstavljaju ancestralni/bazalni tip akustičkog signala kod glavoča u širem smislu (hipoteza prati Malavasi i sur. (2008)), iv. Interspecijski filogenetski odnosi devet vokalnih vrsta glavoča linije Gobius (rodovi Neogobius, Ponticola, Zosterisessor, Gobius, Padogobius), konstruirani na temelju šest kvantitativnih akustičkih parametara, preklapaju se s rezultatima molekularnih analiza, tj. zvukovi prenose filogenetski signal odgovoran za razlikovanje srodnih vrsta.

U ovoj disertaciji, predstavljenoj u četiri znanstvene publikacije, prvi je put istražena, upotrebom tradicionalnih bioakustičkih metoda i analiza, sposobnost stvaranja zvukova i vokalni repertoar (s akustičkim svojstvima) pontokaspijskih glavoča linije Gobius (Gobiiformes; Gobiidae) Neogobius melanostomus, N. fluviatilis i Ponticola kessleri te invazivne vrste rotana Perccottus glenii (Odontobutidae) (publikacije I, II i III). Nadalje, ova

III je disertacija istražila anatomske strukture uključene u proces produkcije zvukova kod rotana, pripadnika sestrinske porodice svih ostalih glavoča u širem smislu (Gobioidei). Rezultati su pokazali kako dosad anatomski ispitani vokalni glavoči (G. paganellus i P. pictus), uključujući i rotana, dijele zajedničke anatomske strukture mehanizma za produkciju zvukova, koji se temelji na kranijalno-pektoralnim mišićima (m. levator pectoralis) i njihovim kontrakcijama (publikacija III). Istodobno, upotrebom molekularnih analiza koje uključuju metode filogenetske rekonstrukcije s više različitih genetičkih biljega (mtDNA i nDNA), rekonstruirani su filogenetski odnosi vokalnih glavoča linije Gobius te je korištenjem komparativne akustičko-molekularne metode istraženo postoji li pozitivna korelacija između genetičke i akustičke divergencije (publikacija IV). Rezultati ove disertacije potvrdili su kako su zvukovi glavoča linije Gobius izrazito specifični za vrstu, dok njihova interspecijska akustička divergencija prati genetičku, što sugerira kako se evolucija zvukova najvjerojatnije mijenjala (i dalje se mijenja) u istom pravcu i jednakom stopom kao i genotipska (neovisno o tipu mtDNA ili nDNA genetičkih biljega) kod navedene skupine riba (publikacija IV).

Znanstveni je doprinos ove disertacije proširivanje znanja o vokalnom repertoaru i akustičkim svojstvima invazivnih pontokaspijskih vrsta rodova Neogobius i Ponticola te rotana P. glenii. Rezultati ove disertacije naglašavaju kako su glavoči izrazito vokalne ribe te kako se, uz standardnu morfološku i molekularnu filogeniju, zvukovi mogu, s velikom pouzdanosti, koristiti u rekonstrukciji filogenetskih odnosa glavoča linije Gobius. Osim toga, ova disertacija naglašava kako se akustička i genetička divergencija znatno preklapaju te slijede vrlo sličan evolucijski obrazac, što potvrđuje važnu ulogu akustičkih signala u divergenciji pojedinih sestrinskih i blisko srodnih europskih glavoča. Ovi rezultati otvaraju potencijalnu mogućnost budućim modernim istraživanjima koja bi se temeljila na automatiziranom prepoznavanju i monitoringu vokalnih glavoča koristeći akustičke baze podataka ili pasivne akustičke snimače. Istraživanje anatomskih struktura uključenih u produkciju zvukova rotana u sklopu ove disertacije naglašava kako je mehanizam za produkciju zvukova, tj. kranijalno-pektoralni mišići, kod glavoča najvjerojatnije evolucijski konzerviran te kako potonji ne koriste druge strukture (npr. plivaći mjehur) u procesu stvaraju zvukova. Stoga ova disertacija znatno doprinosi shvaćanju i proširivanju znanja o načinu na koji glavoči produciraju zvukove tijekom interspecijskih ili intraspecijskih interakcija, što izravno utječe na njihove biološke i ekološke životne karakteristike. Ova disertacija naglašava kako su zvukovi, producirani tijekom reproduktivnih i agresivnih interakcija u sklopu životnog ciklusa, iznimno važan dio komunikacijskog procesa kod glavoča. Zbog toga bi

IV buduća istraživanja blisko povezana s monitoringom ili eradikacijom invazivnih glavoča (kao što su rodovi Neogobius, Ponticola i Perccottus) trebala iskoristiti i implementirati znanja o akustičkoj komunikaciji glavoča radi njihove biokontrole ili potencijalne eradikacije koristeći akustičke zamke/oznake ili druge alate, kao što je već predloženo (Isabella-Valenzi i Higgs, 2016).

THESIS SUMMARY

Gobies (families Gobiidae and Gobionellidae; according to Thacker, 2009) are the most numerous groups of actinopterygian fishes with approximately 1.600 described species, while many undiscovered and cryptic taxa are distributed in almost every aquatic ecosystem (Nelson, 2006; Fricke et al., 2020). Gobies inhabit diverse aquatic habitats, ranging from cold and deep seas, over streams and rivers, to brackish lagoons and estuaries, and in shallow infralittoral they dominate in the abundance and multiformity. Due to their biological characteristics, which include benthic or cryptobenthic lifestyle, speleophylic reproduction and prominent (male) territoriality, gobies are interesting model organisms for laboratory and in situ bioacoustic experiments. Therefore, together with other fish groups (Pomacentridae, Batrachoidiidae, Gadidiae, Serrasalmidae), they are commonly used fish in bioacoustic experiments, with 23 species (~1,2% overall species abundance accrding to Fricke et al., 2020) investigated so far (Zeyl et al., 2016). Until now, most of the scientific attention has been given to sand gobies from Pomatoschisus-lineage (genus Knipowitschia, Economidichthys, Gobiusculus and Pomatoschistus) and Atlantic-Mediterranean gobies from Gobius-lineage (Zosterisessor, Padogobius and Gobius), while the knowledge regarding the acoustic communication, vocal repertoire and acoustic parameters of other interesting groups, like the Ponto-Caspian gobies and Odontobutidae (gobies in a wider sense), are mostly lacking. Multidisciplinary studies indicate that vocal gobies from two different goby families, Gobius paganellus (Gobiidae) and Pomatoschistus pictus (Gobionellidae), posses characteristics and fairly similar body bauplan (structural plan) and anatomical structures in the cranio-pectoral part of their body, therefore emphasizing that gobies (together with other primitive groups of Gobioidei like Odontobutidae) could share similar building elements of their sound production mechanism, making their evolutionary pattern quite recognizable and predictable (Parmentier et al., 2013, 2017). Sound production, outside gobies in the narrow sense (Gobiidae and Gobionellidae), has been investigated in only one additional family named sleepers (Odontobutidae), and the study proposed that Odontobutis obscura could produce pulsatile sounds thanks to rapid pharyngeal teeth collision (Takemura, 1983). Therefore, due to the lack of knowledge about the vocal repertoire and precise anatomical structures involved in sound production in sleepers, a detailed study was necessary to investigate these two aspects of acoustic communication in another sleeper, Perccottus glenii. Furthermore, even though phylogenetic relationships between certain goby groups are still largely undefined, some studies propose that European gobies can be divided into several

VI lineages, namely Gobius- and Aphia- within the subfamily Gobiine; Pomatoschistus- within Gobionelline, according to Agorreta et al. (2013). Molecular analyses, based on the phylogenetic methods with the multilocus approach, indicate that within Gobius-lineage Atlantic-Mediterranean taxa Gobius, Padogobius and Zosterisessor share a close relationship with Ponto-Caspian group (genus Neogobius, Ponticola, Proterorhinus, etc.) widespread in Pontian drainage (Aral, Azov and the Black Sea) and Caspian Sea (Thacker and Roje, 2011; Agorreta et al., 2013). Even though molecular and morphological studies have been conducted in certain Gobius-lineage species, comparative acoustic-genetic research has not been performed for Gobius-lineage gobies or any other teleost group so far. Also, no study has tested empirically and quantitatively the correlation between acoustic diversification and genetic divergence to observe whether they follow the same pattern of evolution, i.e. does the sound carry phylogenetic signal responsible for the vocal taxa discrimination.

This dissertation aimed to broaden the scientific knowledge about the acoustic repertoire (and acoustic parameters), anatomical structures involved in the sound production process and phylogenetic relationships of vocal Gobius-lineage gobies (Gobiiformes, Gobiidae) together with amur sleeper Perccottus glenii (Gobiiformes, Odontobutidae) by using standardized bioacoustic methods and molecular analyses of phylogenetic inference combined with multilocus approach (mtDNA and nDNA). Trough four original scientific publications, this dissertation tried to answer the following questions: i.) How phylogenetically close are Ponto-Caspian gobies with the Atlantic-Mediterranean taxa considering the quantitative and qualitative acoustic parameters, and are only the males vocal sex during intersexual interactions? ii.) Is the sound production mechanism in P. glenii based on the cranio-pectoral muscle contractions and is the pulsatile sound ancestral among gobies in a wider sense (Gobioidei) according to the proposed hypothesis by Malavasi et al. (2008)? iii.) Is there an overlap between the interspecific acoustic divergence of Gobius-lineage gobies (genus Neogobius, Ponticola, Zosterisessor, Gobius, Padogobius), constructed on six quantitative acoustic parameters, with genetic distance, i.e. do sounds in gobies carry phylogenetic signals responsible for the differentiation of investigated vocal gobies?

In this doctoral dissertation, represented by four scientific publications, for the first time by using bioacoustic methods and analyses, the acoustic communication and vocal repertoire of Ponto-Caspian gobies from Gobius-lineage (Gobiiformes; Gobiidae) Neogobius melanostomus, N. fluviatilis and Ponticola kessleri, as well invasive odontobutid Perccottus glenii (Odontobutidae) were tested (publication I, II and III). In addition, this dissertation

VII investigated the anatomical structures closely related to cranio-pectoral girdle in P. glenii which could be responsible for the sound production in this basal Gobioid goby. The results of this dissertation have shown that, together with previously investigated vocal gobies (G. paganellus i P. pictus), P. glenii shares a similar cranio-pectoral structural plan, which is based on levator pectoralis muscle for which it was hypothesized to be responsible for sound production (publication III). By using molecular analyses of phylogenetic inference combined with a multilocus approach (mtDNA and nDNA genetic markers), it was possible to reconstruct the phylogenetic relationships between investigated vocal gobies from Gobius- lineage. In addition, the phylogenetic analysis allowed us to correlate the genetic distance between nine investigated Gobius-lineage gobies with interspecific acoustic divergence (publication IV). The results indicated that the sounds in Gobius-lineage gobies are species- specific, while their acoustic interspecific differentiation follows genotypic divergence (irrespective of genetic marker) which suggests that two traits evolve in the same manner (publication IV).

The scientific contribution of this doctoral dissertation is the broadening of the scientific knowledge about the acoustic communication and vocal repertoire in invasive Ponto-Caspian gobies from genus Neogobius and Ponticola, together with amur sleeper P. glenii. Results of this dissertation strongly suggest that the gobies are highly vocal fish, and together with a standardized method for inferring relationships like morphological or molecular phylogeny, sounds could (as proven by this dissertation) to a certain degree reveal the patter of interspecific divergence between vocal Gobius-lineage gobies. In addition, the observed correlation between acoustic divergence and genetic distance emphasizes that the sounds have a prominent role in the diversification (and possibly speciation) process of closely related Gobius-lineage gobies. These observations present the possible significance of the acoustic signals in indirect and passive usage of the sounds in monitoring or eradication programs of invasive gobies like Neogobius melanostomus, N. fluviatilis, P. kesssleri and P. glenii. Anatomical investigations confirm that gobies in the wider sense, including odontobutid P. glenii, relly on the similar communication pathways, and that the mechanism could be evolutionary conserved (since it is probably based on cranio-pectoral muscle contractions) exploiting already present structures for the sound production through the process of exaptation. Therefore, this dissertation contributes significantly to our understanding of the multimodal communication system in gobies, which also strongly broadens our knowledge about their general biological or ecological life traits. In addition,

VIII since the gobies communicate acoustically during the reproductive seasons, this dissertation emphasizes that the future studies should utilize the information regarding the sound communication in Gobius-lineage gobies (i.e. invasive species from genera Neogobius, Ponticola and odontobutid Perccottus) and implement it in the future monitoring or eradication programs based on acoustic traps/markers like successfully proposed by (Isabella- Valenzi and Higgs, 2016).

1. UVOD

X UVOD UVOD

1.1. Glavoči, najbrojnija skupina riba zrakoperki

Ribe su najbrojnija i najraznolikija skupina kralješnjaka, ponajprije zbog svoje spektakularne morfološke i fenotipske varijabilnosti, ali i opsežnoga geografskog areala te svojih bioloških, često odvedenih, karakteristika (Nelson, 2006, 2016). Kako bismo razumjeli njihovu evoluciju, na početku je potrebno definirati naizgled jednostavan pojam „riba“ i odrediti skup karakteristika koje ih definiraju. Mnogi autori u termin „riba“ često uključuju mnogo nesrodnih skupina kralješnjaka u jedinstvenu neformalnu skupinu. Berra (2001) u svom djelu „Freshwater fish distribution“ iznosi nekoliko kriterija po kojima se može jasnije definirati „riba“, ali sam autor nailazi na mnogo iznimki koje otežavaju naizgled jednostavnu diferencijaciju. Naime, posve logične pretpostavke kako sve ribe žive u vodi, dišu isključivo škrgama, imaju ljuske na površini tijela, pokreću se perajama te su poikilotermni organizmi, ostaju neutemeljene upravo zbog njihove iznimne raznolikosti i plastičnosti, koje pokazuju kako su se ribe danas prilagodile najraznolikijim uvjetima u okolišu te se fenotipski prilično izmijenile (npr. ribe roda Periophthalmus Bloch i Schneider, 1801 mogu preživjeti određeno vrijeme izvan vode; dvodihalice roda Protopterus Owen, 1839 imaju plućno krilo, čime su sličnije tetrapodnim kralješnjacima; somovi i jegulje imaju golu kožu bez ljusaka; ribe roda Apterichtus Duméril, 1806 nemaju peraje; brzi plivači kao što su tune i neki morski psi imaju tjelesnu temperaturu nešto višu od okolnog medija; Berra, 2001). Nelson (2006) pak simplificira pojam te predlaže ovaj opis: „riba je svaki akvatički kralješnjak koji diše škrgama tijekom cijelog života s tjelesnim privjescima u obliku peraja“. Prateći taj kriterij, termin „riba“ predstavlja polifiletsku skupinu kralješnjaka i uključuje, uz hrskavičnjače i ribe zrakoperke i primitivnije oblike kao što su paklare i sljepulje, daleke rođake recentnih „riba“, ali i dvodihalice i latimerije (mesoperke) koje su filogenetski vrlo bliske tetrapodima (Pough i sur., 2009). Kao što je već navedeno, mnogo različitih nesrodnih skupina moglo bi se smatrati „ribama“ u širem smislu. Ipak, pojedini ihtiolozi, tj. stručnjaci koji istražuju ribe, danas termin „riba“ primjenjuju u određenim i specifičnim situacijama, referencirajući se na monofiletsku skupinu zrakoperki – Actinopterygii (Nelson, 2006). Takav je pristup primijenjen i u ovoj disertaciji u kojoj se pojam „riba“ odnosi samo na zrakoperke.

Recentne zrakoperke, u odnosu na druge kralješnjake, dominiraju u gotovo svim vodenim ekosustavima, naseljavajući najraznolikija vodena staništa, od malih potoka na Tibetu na visinama od 5200 m pa do velikih dubina u oceanima ispod 7000 m (Nelson, 2006; Pough i sur., 2009). Neke su zrakoperke čak napustile vodeni medij te mogu preživjeti na kopnu dulje vrijeme, kao što su pripadnici linije Periophthalmus (Nelson, 2006), druge žive u iznimno

1 UVOD ekstremnim staništima (vrsta tilapije Oreochromis alcalicus (Hilgendorf, 1905) živi u jezerima na temperaturi od oko 42 °C, a vrste roda Trematomus Boulenger, 1902 žive na – 2 °C ispod ledenog pokrova na Antartici). Ukratko, zrakoperke imaju karakteristične pokretne i zrakaste peraje (grč. actino = šipčica, ptero = peraja), umjesto tvrdih ili mesnatih peraja, kakve nalazimo kod morskih pasa i mesoperki. Fleksibilne peraje omogućile su im nevjerojatnu preciznost i usmjeren manevar pri pokretanju, hranjenju, reprodukciji, itd. (Pough i sur., 2009). Tijekom svoje evolucijske povijesti zrakoperke su prošle intenzivnu evolucijsku radijaciju i divergenciju u morskim i u slatkim vodama, a danas gotovo da ne postoji vodeno stanište u kojem nije zabilježena barem jedna vrsta zrakoperke (Nelson, 2006, 2016). One su ujedno i najbrojnija skupina kralješnjaka s oko 36.000 recentnih vrsta, što čini više od polovine sveukupno opisanih kralješnjaka (Nelson 2006; Fricke i sur., 2020). Brzina i stopa opisa novih vrsta drastično se mijenjala tijekom godina, na što upućuje činjenica kako je deseto izdanje Systema Naturae (Linnaeus, 1758) navodilo svega 478 vrsta riba. Samo se u 2019. godini otkrila i opisala 351 nova vrsta riba, a prosječan broj iznosi oko 200 opisanih vrsta godišnje, računajući od 2001. godine (Fricke i sur., 2020). Svjetski oceani zauzimaju oko 97% ukupnog volumena vode na Zemlji, dok slatkovodni ekosustavi (ponajprije rijeke i jezera) jedva 0,0093%, a ostatak vode vezan je u ledenjacima, tlu i atmosferi (Horn, 1972). S obzirom na to, zanimljivo je kako je od ukupnog broja opisanih riba njih oko 18.000 isključivo slatkovodno (Fricke i sur., 2020), što potvrđuje kako oko 50% svih riba živi u rijekama i jezerima koji čine samo 0,01% ukupne vode na Zemlji (Horn, 1972). Još zanimljivije zvuči činjenica da prosječno jednu morsku vrstu riba nalazimo u otprilike 113.000 km3 morske vode, a slatkovodnu u samo 15 km3 slatke vode (Horn, 1972). Razlog navedene primijećene diskrepancije u broju slatkovodnih i morskih riba vjerojatno proizlazi iz činjenice kako su ekološke niše u slatkovodnim ekosustavima dugo bile ili djelomično jesu izolirane te su iznimno raznolike (Cohen, 1970). Međutim, uz sve poznato, evolucija i podrijetlo riba i dalje ostaju prilično neistraženo i kompleksno područje znanosti. Zasad je poznato kako najraniji fosili riba koštunjača (Osteichthyes) datiraju još iz kasnog silura (oko 420 milijuna godina, https://www.britannica.com/animal/fish; Pough i sur., 2009), zrakoperki iz devona (oko 400 milijuna godina, Pough i sur., 2009), dok se tijekom razdoblja krede (mezozoik; 144 – 66 milijuna godina) pojavljuju njihovi recentni oblici (Near i sur., 2013). Tijekom kenozoika dolazi do ubrzane evolucije i znatne raznolikosti zrakoperki (Near i sur., 2013), koja je vjerojatno bila omogućena znatnim promjenama i specijalizacijama koštanih dijelova lubanje i aparata za hranjenje (uključujući čeljusti), zajedno s razvojem pokretljivih i fleksibilnih peraja (Pough i sur., 2009). Zrakoperke se danas koriste u znanstvenim

2 UVOD istraživanjima kao modelni organizmi u raznim ekološkim, etološkim, evolucijskim i biogeografskim studijama, a od neizmjerne su važnosti za ljude i kao važan izvor hrane i proteina širom svijeta od početka ljudske povijesti (Nelson, 2006). Zbog svoje uloge u ekosustavu i važnosti za ljude dio su internacionalnih sporazuma i dogovora, koriste se kao indikatori onečišćenja i narušenosti vodenih ekosustava, a imaju i izraženu rekreacijsku te ekonomsku ulogu (Nelson, 2006).

Glavoči (u užem smislu, pripadnici porodica Gobiidae i Gobionellidae, filogenija prema Thacker, 2009) predstavljaju jednu od najraznolikijih skupina riba unutar reda Gobiiformes (Thacker, 2009; Bentacur i sur., 2013), ali i jednu od najstarijih skupina zrakoperki budući da su se pojavili u eocenu (prije 51 – 55 milijuna godina; Near i sur., 2013; Thacker, 2014). Među ostalim, oni su i najbrojnija skupina riba općenito, s gotovo 1250 validnih vrsta opisanih do danas, iako njihov broj izrazito varira od autora do autora te se nove vrste svakodnevno opisuju (Berra, 2001; Nelson, 2006; Thacker, 2009; Near i sur, 2013; Fricke i sur., 2020). Široko su rasprostranjeni (Slika 1), a najčešće naseljavaju plitka morska staništa (infralitoral) gdje su najbrojnija skupina riba (Nelson, 2006), bočate estuarije ili lagune te slatkovodne rijeke, jezera i krške potoke u umjerenim, tropskim i suptropskim područjima, s posebnim naglaskom na tropski indopacifički pojas gdje su glavoči izrazito brojni (Berra, 2001; Nelson, 2006; Thacker, 2009, 2015). Zanimljivo je kako nekoliko vrsta naseljava naizgled nepovoljna vodena staništa kao što su hladna mora oko Norveške i Islanda te subpolarne potoke u Sibiru (Berra, 2001; Kovačić i Svensen, 2019). Na koraljnim grebenima glavoči su jedna od najbrojnijih skupina riba i čine oko 20% ukupne raznolikosti vrsta i 35% ukupnog broja riba (Winterbottom i sur., 2011).

Prema načinu života glavoči zauzimaju gotovo sve ekološke niše u vodenom mediju. Oni su većinom pridnene ili bentičke ribe koje su svojim navikama blisko vezane za plitko morsko ili slatkovodno dno (većina vrsta živi na dubinama od 2 do 10 metara) iako postoji i nekoliko pelagičkih oblika (npr. Aphia minuta (Risso, 1810), Crystallogobius linearis (Düben, 1845), Pseudaphya ferreri (Buen i Fage, 1908) (Kovačić i Patzner, 2011)), dok su neke i batofilne vrste, kao što je primjerice Gobius roulei, vrsta zabilježena na dubinama od oko 400 metara (Kovačić i Patzner, 2011). Neke vrste, kao što su glavoči linije Periophthalmus (rodovi Boleophthalmus Valenciennes, 1837, Periophthalmus, Periophthalmodon Bleeker, 1874 i Scartelaos Hamilton, 1822) prilagodile su se amfibiotskom načinu života, tj. izlaze iz vode u potrazi za hranom, kreću se uz pomoć prsnih peraja po plitkom muljevitom dnu te se nakon nekog vremena provedenog vani vraćaju u svoja muljevita vodena staništa (Nursall, 1981;

3 UVOD

Berra, 2001; Nelson, 2006). Zanimljiv primjer su i slatkovodni glavoči roda Clamydogobius Whitley, 1930, koji naseljavaju arteške izvore i lokve u pustinjama središnje Australije te predstavljaju odvojenu slatkovodnu skupinu u odnosu na svoje morske pretke (Larson, 1995). Pojedine vrste žive u simbiotskom odnosu s drugim životinjama (npr. spužve, rakušci, ježinci; najpoznatiji primjer vrsta Crytocentrus steinitzni (Klausewitz, 1974) i dekapodni rakušac roda Alpheus Fabricius, 1798 (Preston, 1978; Karplus, 1979)), dok su vrste roda Gobiosoma Girard, 1858 „čistači“ te se hrane ektoparazitima drugih tropskih riba na koraljnim grebenima (Limbaugh, 1961; Losey, 1987; Nelson, 2006). Glavoče se većinom može naći na mekanom sedimentnom supstratu kao što su muljevita ili pjeskovita dna, dok znatan broj vrsta živi i na tvrdom supstratu (kamenje, stijene, antropogeni materijal) te u špiljama (Kovačić i Patzner, 2011). Također, neke vrste su kriptobentičke (tj. žive u pukotinama, šupljinama ili ispod raznog materijala, prema Miller, 1979, 1996), što otežava njihovo istraživanje in situ (Kovačić i Patzner, 2011). Fenotipski, glavoči su izrazito raznolika skupina, no sve ih karakterizira nekoliko zajedničkih morfoloških i bioloških osobina. Glavoči su male ribe čija prosječna duljina tijela rijetko prelazi više od 20 centimetara (cm) ukupne duljine (TL), dok većina vrsta doseže duljine od 5 do 10 cm (najmanji glavoč na svijetu je s Filipina Pandaka pygmaea Herre, 1927, čija ukupna duljina ne prelazi 1,3 cm, dok je najveći glavoč Gobioides broussenetii Lacepède, 1800 iz Karipskog mora, koji doseže veličine do 50 cm; Miller, 1986; Berra, 2001; Kovačić i Patzner, 2011). Većina glavoča nema plivaći mjehur (tj. hidrostatski organ za održavanje u vodenom stupcu povezan s probavnim sustavom) i kompletno razvijen sustav bočne pruge na tijelu (skupina osjetilnih tjelešaca koja registriraju pokrete i pritisak uz bočne strane tijela ribe), dok su navedene strukture prisutne kod većine današnjih riba (Berra, 2001; Nelson, 2006; Kottelat i Freyhof, 2007). Smatra se kako je nedostatak plivaćeg mjehura prilagodba na bentički način života, dok je bočna pruga djelomično reducirana i ograničena samo na glavu u obliku kanala i vanjskih nizova osjetilnih papila, koji čine važan skup dijagnostičkih morfoloških karakteristika za razlikovanje vrsta i rodova (Miller, 1986). Nadalje, gotovo svi glavoči imaju adhezivni (trbušni) disk nastao spajanjem trbušnih (pterygia ventralia) peraja, a smatra se kako im on, zbog podtlaka koji stvara, omogućuje olakšano održavanje u pridnenim dijelovima (Berra, 2001; Nelson, 2006). Tijekom reproduktivne sezone glavoči su izrazito teritorijalne ribe koje prezentiraju stereotipno ritualno udvaranje i intraseksualnu kompeticiju među mužjacima za ženku i gnijezdo (ljušture školjkaša, šupljine, pukotine, antropogeni materijal i dr.), odakle privlače partnerice na mrijest (Lindström i Hellström, 1993; Malavasi, 2003, 2009). Kod većine vrsta ženka polaže jaja na gornju površinu gnijezda (tzv. speleofilna reproduktivna taktika), a mužjak ga priprema i čuva

4 UVOD od drugih potencijalnih udvarača te se brine o mladima (roditeljska briga) dok se ne izlegnu iz jaja (Tavolga, 1956; Torricelli i sur., 1985; Lindström i Hellström, 1993).

Pokazalo se kako se današnji žarišni centar raznolikosti (eng. hot-spot) glavoča nalazi u Indo-australskom području (plitka tropska područja, tj. otočne skupine koje se pružaju u smjeru juga od Filipina, preko Indonezije do Nove Gvineje) iako je nekoliko skupina prisutno u Mediteranskom moru, Pacifičkom oceanu (istočni dio), Karipskom moru te u Atlantskom oceanu (istočno i zapadni dio) (Thacker, 2015). Fosilni nalazi pokazuju kako su se pripadnici podreda Gobioidei (glavoči u širem smislu, porodice Odontobutidae, Ryachthydidiae, Eleotridide, Butidae, Gobionellidae i Gobiidae, prema Thacker, 2009) razvili u slatkim, bočatim i rubnim morskim staništima tijekom kenozoika, s nekoliko primjeraka fosila poznatih iz eocena, oligocena i miocena u Europi (Gierl i sur., 2013; Prikryl, 2014). U mediteranskom području glavoči (Gobiidae i Gobionellidae) su najbrojnija skupina riba s oko 60 zabilježenih vrsta (Kovačić, 2020), dok u crnomorskom području obitavaju 33 vrste, a u sjeveroistočnom (SI) dijelu Atlantskog oceana 47 vrsta (za sve reference, vidi Kovačić i Patzner, 2011). Filogenija i sistematika glavoča i dalje ostaje djelomično nerazriješena, uz prisutnost mnogih fenotipski (morfološki) izrazito sličnih vrsta (od kojih neke još uvijek čekaju determinaciju i znanstveni opis) koje dodatno otežavaju ionako detaljne i kompleksne analize (Kovačić i Patzner, 2011). Poznato je kako glavoči u širem mediteranskom području (Mediteransko i Crno more te SI Atlantik) naseljavaju najraznolikija morska staništa, od plitkih obalnih ekosustava do velikih dubina preko 300 m (Miller, 1986). Od ukupnog broja, 27% vrsta naseljava vodeni stupac od površine do 2 m dubine, 31% naseljava pojas od 2 do 10 m, a samo 13% vrsta živi na dubinama većim od 50 m (Kovačić i Patzner, 2011). Većinom su bentičke ribe blisko vezane uz dno, pri čemu 60% vrsta preferira mekana dna (pijesak, mulj, sitni šljunak) u odnosu na tvrdi supstrat ili špilje (Kovačić i Patzner, 2011). Jedna je od najmanjih vrsta navedenog područja Speleogobius trigloides Zander i Jelinek, 1976 (do 2.26 cm TL), a jedna od najvećih Gobius cobitis Pallas, 1814 (do 27 cm TL), pri čemu samo 34% vrsta naraste iznad 10 cm TL (Miller, 1986; Kovačić i Patzner, 2011). Po brojnosti i ukupnoj raznolikosti vrsta sjeverni dio Mediteranskog mora (zajedno s Jadranskim morem; broj vrsta iznosi između 46 i 43 vrste) i Levantsko područje (istočni Mediteran s 29 do 37 vrsta) razlikuju se od središnjeg i zapadnog Mediterana (do 12 vrsta), dok se Crno more izdvaja zbog specifične zajednice i sastava vrsta u odnosu na druga područja (33 vrste). U SI Atlantiku živi manji broj vrsta u odnosu na Atlantski obalni dio (Kovačić i Patzner, 2011). Kovačić i Patzner (2011) navode kako je upravo sjeverni Mediteran najbrojniji vrstama i

5 UVOD mediteranskim endemima, a 2/3 (14 vrsta) svih mediteranskih endema egzistira u njegovim sjeverozapadnim dijelovima i Egejskom, Jonskom i Jadranskom moru s naglaskom na Jadran koji s 22 mediteranske endemske (morske) vrste dominira u odnosu na ostala područja. Zapažena brojnost i bogatstvo vrsta u Mediteranu posljedica je najvjerojatnije geoloških, hidroloških i klimatskih događaja na tom području, na što dodatno upućuje i broj endemskih vrsta (samo šest u SI Atlantiku nasuprot 24 u Mediteranu). Pretpostavlja se kako su današnje mediteranske endemske vrste vjerojatno evoluirale od svojih predaka iz Atlantskog oceana koji su nakon mesinske krize saliniteta (MSC) za vrijeme miocena (presušivanje Mediteranskog mora u razdoblju 5,9 ̶ 5,3 milijuna godina; Hsü i sur., 1977; Rögl, 1998; Krijgsman i sur., 1999; Quinard i Tomasini, 2000) mogle migrirati kroz Gibraltarski tjesnac te proći u Mediteransko more, koje je poslužilo kao refugij toplovodnim vrstama tijekom glacijalnih razdoblja (Briggs, 1974; Almada i sur., 2001). Među ostalim, temperaturne fluktuacije tijekom glacijalnih i interglacijalnih razdoblja u pleistocenu (2,5 ̶ 0,011 milijuna godina) mogle su utjecati na diversifikaciju i evoluciju tadašnje faune riba mediteranskog područja (uključujući i glavoča) zahvaljujući ponovnoj invaziji i rekolonizaciji Mediterana atlantskim vrstama (Almada i sur., 2001). U odnosu na morske vrste, mediteransko područje (uključujući Crno more i Kaspijsko jezero) naseljava 50 vrsta slatkovodnih glavoča zabilježenih u rijekama, potocima, jezerima i izvorima, sa slijevovima Jadranskog i Egejskog mora (od sjeverne Italije, južno preko Grčke do zapadne Anatolije) kao dominantnim žarišnim centrima njihove raznolikosti (Miller, 2003, 2004; Šanda i Kovačić, 2009; Freyhof, 2011). Zbog svoje raznolikosti i širokog areala pojedini autori (Miller, 1986, 2003, 2004; Smirnov, 1986) smatraju kako u europskim vodotocima i širem mediteranskom području (uključujući SI Atlantik, Mediteransko more i Crno more) postoje dvije skupine glavoča s alopatrijskom (diskontinuiranom) distribucijom: atlantsko-mediteranska (SI Atlantik i Mediteransko more) i pontokaspijska (Crno, Aralsko i Azovsko more i Kaspijsko jezero s pritocima). Atlantsko- mediteranska skupina predstavlja skupinu rodova (za ovu disertaciju najbitniji Gobius Linnaeus, 1758 i Zosterisessor Whitley, 1935, koji se danas smatra mlađim sinonimom roda Gobius), čiji su filogenetski odnosi i sistematski položaj kompleksniji i zamršeniji u odnosu na pontokaspijske vrste. Pontokaspijske vrste su monofiletska skupina glavoča sastavljena od vrsta rodova Babka Iljin, 1927, Benthophiloides Beling i Iljin, 1927, Benthophilus Eichwald, 1831, Caspiosoma Iljin, 1927, Mesogobius Bleeker, 1874, Neogobius Iljin, 1927, Ponticola Iljin, 1927, Proterorhinus Smitt, 1900 (Miller, 2003, 2004; Neilson i Stepien, 2009a; Medvedev, 2013), od kojih su neke vrste alohtone i invazivne u europskim vodotocima i sjevernoj Americi (npr. Neogobius melanostomus (Pallas, 1814), N. fluviatilis (Pallas, 1814),

6 UVOD

Ponticola kessleri (Günther, 1861) (Charlebois i sur., 1997; Stepien i Tumeo, 2006; Neilson i Stepien, 2009a)). One su, zajedno s mediteranskim Padogobius Berg, 1932 rodom, pleizomorfne s eurihalinim mediteranskim rodom Zosterisessor u redukciji ili gubitku slobodnih šipčica prsnih peraja (Miller, 2004), dok, kao i P. nigricans (Canestrini, 1867), imaju iznimno povećan broj kralješaka (29 ̶ 35) i reduciran plivaći mjehur u odnosu na mediteranske rodove (Miller, 2003, 2004).

Slika 1. Rasprostranjenost glavoča u slatkovodnim (crno) i morskim (sivo) ekosustavima diljem svijeta (preuzeto iz Berra, 2001). Napomena: autor uključuje sve glavoče u užem smislu unutar porodice Gobiidae (Gobiiformes; Gobioidei), a oni su se poslije modernim molekularnim analizama razdvojili na dvije zasebne porodice (Gobiidae i Gobionellidae prema Thacker, 2009, 2015).

1.2. Pregled najvažnijih filogenetskih istraživanja europskih glavoča linije Gobius (Gobiiformes, Gobiidae)

Iako su iznimno brojna skupina riba, glavoči se često izostavljaju iz evolucijskih i biogeografskih studija (Bellwood i Wainwright, 2002; Cowman, 2014), ponajprije zbog iznimne, ali i kompleksne raznolikosti (fenotipska, morfološka, ekološka, etološka, itd.), zamršene taksonomije (pogotovo bliskih vrsta, od kojih neke još uvijek čekaju na svoj opis), male veličine tijele te skrovite naravi, tj. kriptobentički način života (Thacker, 2011, 2015). Razvojem novih molekularnih metoda, u odnosu na tradicionalne morfološke i

7 UVOD morfometrijske analize, kompleksna slagalica povezana s evolucijom mnogih vrsta i njihovim filogenetskim odnosima počela se djelomično sastavljati (Thacker, 2009). Donedavno su glavoči smatrani podredom unutar grgečki (Gobioidei; Perciformes), te su zajedno s podredovima Percoidei i Labroidei činili 3/4 ukupne brojnosti svih opisanih grgečki (Berra, 2001; Nelson, 2006). U ovoj disertaciji pratit će se filogenetski odnosi i taksonomski položaj glavoča koje predlažu Thacker (2009) i Betancur-R i sur. (2017), prema kojima se glavoči izdvajaju od grgečki, a red Gobiiformes postaje dio subdivizije Percomorphaceae (serija Gobiaria, Betancur-R i sur., 2017) unutar kojega se prepoznaje podred Gobioidei sa šest porodica: Rhyacichthyidae, Odontobutidae, Butidae, Eleotridae, Gobionellidae i Gobiidae (Thacker, 2009). Također, Agorreta i sur. (2013) grupiraju sve glavoče unutar jedne porodice Gobiidae, ali prepoznaju unutar nje nekoliko nezavisnih evolucijskih linija, od kojih je za ovu disertaciju najvažnija linija Gobius. Treba imati na umu kako se, prema Thackeru (2009), glavoči dijele na dvije porodice (Gobiidae i Gobionellidae), dok Agorreta i sur. (2013) prepoznaju samo jednu porodicu (Gobiidae) s nekoliko potporodica, a za nas su najvažnije Gobiinae i Gobionellinae. Kao što je već navedeno, u ovoj disertaciji slijedit će se sistematika porodica koju predlaže Thacker (2009), dok će se nazivi linija preuzeti od Agorreta i sur. (2013).

Jedno od prvih interspecijskih istraživanja filogenetskih odnosa glavoča mediteranskog područja objavili su Penzo i sur. (1998). Cilj je tog istraživanja bio analizirati sekvence 12S i 16 S rRNA mitohondrijskih gena za trinaest vrsta glavoča (rodovi Gobius, Padogobius, Pomatoschistus Gill, 1863, Knipowitschia Iljin, 1927, Zebrus de Buen, 1930, Zosterisessor) kako bi se istražili odnosi među pojedinim vrstama te kako bi se razriješilo pitanje kada se prvi put javlja preferencija na slatkovodno stanište kod proučavanih vrsta (tri su vrste u analizi bile isključivo slatkovodne), uzimajući u obzir geološki kontekst šireg mediteranskog područja. Prijašnja su istraživanja navodila kako se preferencija na slatkovodna staništa razvila nekoliko puta nezavisno, upućujući na polifiletske odnose među talijanskim glavočima (Miller i sur., 1994). Metodama najveće parsimonije (MP) i metodom genetske udaljenosti (NJ), rezultati istraživanja pokazali su kako su se istraživane vrste grupirale u dvije grupe (eng. clade) na filogenetskom stablu: prva obuhvaćajući vrste iz rodova Knipowitschia i Pomatoschistus, a druga grupirajući sve ostale vrste (rodovi Gobius, Padogobius, Zebrus, Zosterisessor). U prvoj su se grupi oba roda Knipowitschia i Pomatoschistus pokazala parafiletskima, kao i rod Gobius iz druge grupe jer su se sve morske vrste s preferencijom na muljevito/pjeskovito stanište (Gobius niger Linnaeus, 1758, G. auratus Risso, 1810, G.

8 UVOD bucchichi Steindachner, 1870, i Zosterisessor ophiocephalus (Pallas, 1814)) grupirale zajedno. Također, u drugoj grupi rod Padogobius pokazao se parafiletskim zato što se P. nigricans grupirao kao sestrinska vrsta s G. paganellus Linnaeus, 1758, dok je P. bonelli (Bonaparte, 1846) prikazan bazalnom vrstom u odnosu na prijašnje dvije. Zebrus zebrus (Risso, 1827) sestrinska je vrsta u odnosu na G. paganellus + Padogobius spp. grupu (Penzo i sur., 1998). Dobiveni rezultati pokazali su kako se preferencija na slatkovodno stanište talijanskih glavoča (Orsinigobius punctatissimus (Canestrini, 1864), Padogobius nigricans i P. bonelli) razvila nekoliko puta (polifiletsko podrijetlo) iako je odnos između P. nigricans i G. paganellus ostao nerazriješen, odnosno iako morfološki slični, P. nigricans i P. bonelli pokazali su se kao nesestrinske vrste (Penzo i sur., 1998). Ti su rezultati pokazali kako je preferencija na slatku vodu najvjerojatnije posljedica jednog od dvaju scenarija: preferencija se javlja nekoliko puta nezavisno kod vrsta roda Padogobius ili samo jednom, kod pretka grupe G. paganellus + Padogobius spp., nakon čega se sekundarno gubi kod G. paganellus koji je eurihalina/morska vrsta (Penzo i sur., 1998). Stoga je bliska filogenetska srodnost Z. ophiocephalus s drugim vrstama roda Gobius, zajedno s navedenim slučajem G. paganellus + Padogobius spp., pokazala kako je rod Gobius najvjerojatnije parafiletskog podrijetla (McKay i Miller, 1991; Penzo i sur., 1998). Autori su zaključili kako je inicijalno razdvajanje dviju grupa (glavoči pijeska i Gobius, + Padogobius + Zosterisessor) vjerojatno rezultat odvajanja Parateisa (Pontokaspijsko područje) od mediteranskog oceana Tetisa prije 10 ̶ 12 milijuna godina (Steininger i Rögl, 1984). Prilagodba istraživanih glavoča na slatkovodno stanište rezultat je vjerojatno mesinske krize saliniteta u Mediteranu (prije 5,5 milijuna godina) i razdoblja poznatog kao „Lago Mare“ (Hsü i sur., 1977) ili zapadnog širenja azijske ihtiofaune kroz Panonski kanal prije 24 milijuna godina u miocenu (Banarescu, 1992).

Radi rješavanja filogenetskih odnosa i sistematskog položaja vrsta iz podreda Gobioidei, Thacker (2009) provodi detaljno molekularno istraživanje i analizu DNK sekvenci četiri mitohondrijska biljega (COI, ND1, ND2 i cyt b) izoliranih iz jedinki glavoča reda Gobiiformes, a u analizu uključuje podatke o nekoliko njihovih fenotipskih i ekoloških karakteristika. Rezultati tog istraživanja pružili su dublji uvid u zamršenu sistematiku glavoča, naglašavajući kako bi se redu Gobiiformes trebala vratiti taksonomska jednakost u odnosu na Perciformes (tj. glavoče odvojiti od grgečki u zasebnu taksonomsku cjelinu, tj. red) te kako unutar navedenog reda postoje tri odvojene grane (podreda), od kojih monofiletski podred Gobioidei obuhvaća šest porodica (Odontobutidae, Rhyacichthyidae, Eleotridae, Butidae, Gobiidae i Gobionellidae) u odnosu na prije predloženih devet (Miller, 1973; Springer, 1983;

9 UVOD

Hoese, 1984; Hoese i Gill, 1993; Johnson i Brothers, 1993; Thacker, 2000). Provedeno je istraživanje doprinijelo sveukupnom znanju o filogenetskim odnosima pojedinih skupina glavoča te simplificiralo zamršenu sistematiku grupirajući neke manje skupine (potporodice) u više taksonomske cjeline (porodice). Uz to, rezultati su pokazali kako su dvije najbrojnije i najraznolikije skupine unutar Gobioidei upravo Gobiidae i Gobionellidae (glavoči u užem smislu), od kojih brojnija Gobiidae obuhvaća većinom morske vrste, a Gobionellidae slatkovodne i bočate taksone (Thacker, 2009).

Neilson i Stepien (2009a) istražili su interspecijske filogenetske odnose i sistematski položaj pontokaspijskih vrsta (skupina Benthophilinae Iljin 1927) koristeći analizu DNK sekvenci za dva mitohondrijska (cyt b i COI) i dva jezgrena (RAG1 i S7) gena uz pomoć metoda filogenetske rekonstrukcije: metoda najveće parsimonije (MP), metoda najveće vjerodostojnosti (ML) i Bayesova metoda (BI). Dodatni su ciljevi istraživanja bili istražiti monofiletsko podrijetlo pontokaspijskih vrsta, objasniti ekološku prirodu njihova invazivnog karaktera te uklopiti hipotezu o njihovoj evoluciji u geološke događaje i paleoklimatske promjene navedenog područja (Aralsko, Azovsko i Crno more te Kaspijsko jezero s pritocima, prema Reid i Orlova, 2002). Istodobno, istraživanje je provedeno kako bi se pontokaspijski glavoči, prije svrstani u jedinstveni rod Neogobius sensu Berg (1949), razdvojili, prateći taksonomske norme, u odgovarajuće sistematske kategorije (rodove i vrste). Kombinirani filogenetski rezultati, temeljeni na četiri gena, pokazali su kako su pontokaspijski glavoči zaista monofiletska skupina unutar porodice Gobiidae, unutar kojih se razdvajaju tri skupine visoke podržanosti (>85%): Neogobiini (vrste roda Neogobius), Benthophilini (rodovi Benthophilus i Caspiosoma) te Ponticolini (Proterorhinus, Mesogobius, Babka i Ponticola). Na filogenetskom stablu skupina Neogobiini pokazala se sestrinskom grupom drugoj grupi sastavljenoj od vrsta grupe Benthophilini i Ponticolini (95% ML bootstrap). Također, detaljna je analiza divergencije (na cyt b koristeći ML) pokazala kako su se pontokaspijski glavoči odvojili od ostatka Gobiidae prije oko 39 milijuna godina, s naglaskom na tome kako se posljednji zajednički predak vrsta rodova Neogobius, Babka, Proterorhinus i Ponticola razdvojio prije oko 10 milijuna godina kada su se razdvojili Pontokaspijski i Panonski bazeni (Reid i Orlova, 2002; Neilson i Stepien, 2009a). Razdvajanje Crnog mora od Kaspijskog jezera (prije 5,8 – 5 milijuna godina) u isto vrijeme kao i mesinska kriza saliniteta u Paleomediteranu i crnomorskom slijevu, dovelo je do razdvajanja vrsta unutar skupine Neogobiini. Rezultati molekularne analize navode kako su se nakon 5,0 milijuna godina sve skupine Benthophilinae najvjerojatnije razdvojile, nakon čega

10 UVOD je počela interspecijska evolucija pojedinih rodova, i morskih i slatkovodnih (Reid i Orlova, 2002; Neilson i Stepien, 2009a).

Thacker i Roje (2011) analizirali su, kao dopunu prijašnjem istraživanju (Thacker, 2009), izolirani set DNK sekvenci za pet gena (tri mtDNA – CO1, ND1 i ND2 i dva nDNA – RAG2 i Rho) za 100 vrsta iz porodice Gobiidae (potporodice Gobiinae, Microdesmidae, Ptereleotridae, Kraemeriidae i Schindleriidae) radi dobivanja boljeg uvida u njihove filogenetske odnose. Provedeno istraživanje uključilo je u molekularne analize genski materijal gotovo svih poznatih grupa Gobiidae, uzimajući u obzir i američke, pontokaspijske, atlanstko-mediteranske vrste. Rezultati provedenog istraživanja pokazali su kako unutar porodice Gobiidae postoje tri monofiletske grupe (eng. clades) i trinaest linija (eng. lineages), čiji su međusobni odnosi na filogenetskom stablu slabo poduprti, ali je zato unutar linija poduprtost grananja iznosila 80% (boostrap) ili više (Thacker i Roje, 2011). Zanimljivo je kako su se unutar treće i ujedno najveće linije grupirali rodovi Novog svijeta i Indonezije, ali i atlantski, pontokaspijski i mediteranski rodovi (Thacker i Roje, 2011). Detaljniji je uvid u filogenetske odnose otkrio kako na filogenetskom stablu pontokaspijski glavoči čine sestrinsku grupu s atlantsko-mediteranskim vrstama (rod Gobius i Zosterisessor) s čvorom visoke podržanosti (80%), dok afrički rodovi Coryogalops Smith, 1958 i Caffrogobius Smitt, 1900 čine sestrinsku grupu s već navedenom skupinom (Thacker i Roje, 2011). Autori naglašavaju kako navedena grupa (pontokaspijske i atlantsko-mediteranske vrste) čini monofiletsku skupinu, a rezultati molekularnog istraživanja koje su proveli Neilson i Stepien (2009a) uputili su na sličan scenarij, s odnosima među pojedinim pontokaspijskim vrstama malo drugačijim u odnosu na Thacker i Roje (2011) (unutar skupine Benthophilinae Neogobiini čini bazalnu skupinu, dok su Ponticolini i Benthophilini sestrinske skupine). Ti su rezultati objašnjeni u svjetlu geološke prošlosti i događaja šireg mediteranskog područja te naglašavaju kako su mesinska kriza saliniteta (tj. isušivanje mediteranskog područja i formiranje slanih i slatkih jezera diljem mediteranskog i pontokaspijskog područja prije 5,9 – 5,3 milijuna godina) i naknadan dotok morske vode kroz Gibraltarski tjesnac (prije 5,3 milijuna godina) imali znatan utjecaj na diversifikaciju vrsta navedenog područja (Penzo i sur., 1998; Simonović, 1999; Neilson i Stepien, 2009a).

Nadalje, Agorreta i sur. (2013) analizirali su sekvence pet genetičkih biljega (dva mtDNA – 12S/16S i cyt b - i tri nDNA – zic1, sreb2, RAG1) od 222 vrste podreda Gobioidei (glavoči u širem smislu) te pokušali detaljnije rekonstruirati njihove filogenetske odnose koristeći ML i BI metode, s posebnim naglaskom na europske vrste i filogenetske odnose među njima.

11 UVOD

Slično kao i Thacker (2009), rezultati molekularnog istraživanja Agorreta i sur. (2013) pokazali su kako je podred Gobioidei monofiletskog podrijetla te kako stabla upućuju na slične topologije, tj. filogenetski najbazalnije porodice među Gobioidei su upravo Rhyacichthyidae i Odontobutidae koje na filogenetskom stablu predstavljaju sestrinsku grupu za sve ostale porodice unutar Gobioidei (Slika 2). Naknadno se na filogenetskom stablu nakon Rhyacichthyidae i Odontobutidae odvojila porodica Eleotridae, a nakon nje Butidae koja se pokazala sestrinskom grupom porodici Gobiidae, unutar koje su autori prepoznali dvije visokopodržane podgrupe/potporodice (Gobiinae i Gobionellinae, Agorreta i sur., 2013). Uz to, rezultati navedenog istraživanja detaljnije su istražili filogenetske odnose među pojedinim europskim skupinama glavoča (36 rodova) koji se, s obzirom na molekularne analize temeljene na pet gena, grupiraju u tri odvojena evolucijska klastera: liniju Pomatoschistus, liniju Aphia Risso, 1827, i liniju Gobius (prva je dio podgrupe Gobionellinae, dok su zadnje dvije dio potporodice Gobiinae, Agorrera i sur., 2013). Većina se europskih glavoča, zajedno s afričkim skupinama (Sufflogobius Smith, 1956, Gorogobius Miller, 1978, Corcyrogobius Miller, 1972, Chromogobius Buen, 1930 i dr.), grupirala u liniju Gobius, što potvrđuje kako nekoliko istočno atlantskih tropskih rodova dijeli bliske filogenetske odnose s europskim vrstama, ali i da afrički glavoči prema Thacker i Roje (2011) ne predstavljaju monofiletsku skupinu (Agorreta i sur., 2013). Kao i u prijašnjim analizama (Neilson i Stepien, 2009a; Thacker i Roje, 2011, Tornabene i sur., 2013), pontokaspijski glavoči pokazali su se, na temelju provedenih analiza, monofiletskom skupinom blisko povezanom s europskim rodovima Gobius i Zosterisessor (Agorreta i sur., 2013). Također, rezultati provedenog istraživanja pokazali su kako je linija Gobius, uklanjanjem enigmatičnog taksona Kraemeria Steindachner, 1906 iz analiza, blisko povezana s rodom Priolepis Valenciennes, 1837 rodom (Pacifički ocean) i skupinom Gobiosomatini (Karipsko more) (Agorreta i sur., 2013).

12 UVOD

Slika 2. Filogenetsko stablo (filogram) vrsta unutar podreda Gobioidei rekonstruirano na temelju pet različitih genetičkih biljega (dva mitohondrijska i tri jezgrena) metodom najveće vjerodostojnosti (ML). Na stablu se odvaja pet porodica unutar podreda Gobioidei (Odontobutidae, Rhyacichthyidae, Eleotridae, Butidae, Gobiidae), dok su glavoči u užem smislu podijeljeni na dvije potporodice unutar Gobiidae (Gobiinae i Gobionellinae). Ispunjeni čvorovi na podržanost ML (BP > 70%), dok neispunjeni čvorovi na podržanost ML (BP < 70%). Skala predstavlja supstitucije po mjestu na grani stabla. Preuzeto iz Agorreta i sur., 2013.

Biogegrafske karakteristike te migracijske rute glavoča tijekom kenozoika (Gobiiformes, Gobioidei) istražene su u Thacker (2015) filogenetskim istraživanjem mitohondrijskih ND1, ND2, COI i niklearnih RNF, Rho, RAG2 gena, za ukupno 83 glavoča u užem smislu (Gobiidae i Gobionellidae) s Oxyeleotris lineolata (Steindachner, 1867) (Butidae) kao vanjskom grupom (eng. outgroup), radi temporalnog (vremenskog) kalibriranja pojedinih grupa i određivanja njihova ancestralnog areala. Također, u istraživanju su molekularnim podacima integrirani i podaci o recentnoj distribuciji i fosilnim nalazima glavoča kako bi se

13 UVOD rekonstruirali povijesni pravci i rute invazije pojedinih grupa i rodova glavoča. Prema literaturi, tijekom kenozoika postojalo je nekoliko vrućih točaka povezanih s uzajamnom rapidnom diversifikacijom i radijacijom koralja i glavoča (zapadni Tetis – Mediteran, zapadni Indijski ocean, Indopacifik) (Wood, 1999; Bellwood i Wainwright, 2002; Renema i sur., 2008). Dobiveni rezultati molekularnog istraživanja Gobiidae i Gobionellidae, s topologijama na filogenetskom stablu i odnosima među pojedinim grupama, nisu znatno odstupali od prijašnjih istraživanja (Thacker i Roje, 2011; Thacker, 2013, 2014; Agorreta i sur., 2013), a filogenetsko je istraživanje pokazalo kako je linija Gobius (mediteranski i pontokaspijski glavoči) blisko povezana (sestrinski položaj) s američkom skupinom Gobiosomatini. Vremenski kalibrirana filogenija pokazala je kako su glavoči porodica Gobiidae i Gobionellidae najvjerojatnije evoluirali u razdoblju od oko 12 – 13 milijuna godina tijekom eocena, od otprilike 48,7 pa do 36,2 milijuna godina, s ancestralnin arealom za Gobiidae + Gobionellidae smještenim u Indopacifiku (gdje su i danas najbrojnija skupina), a ne u Mediteranu ili zapadnom Indijskom oceanu (Thacker, 2015). Prema rezultatima istraživanja, smatra se kako je za skupine unutar Gobionellidae evolucijsko podrijetlo vezano za Mediteransko more, dok je linija Pomatoschistus do danas zadržala takvu rasprostranjenost. Za skupine Gobiidae, ancestralnim arealom predložen je Indopacifik, dok je linija Gobius najvjerojatnije evoluirala na području zapadnog Indijskog oceana (Thacker, 2015; Slika 3). Nakon evolucije u Indopacifiku glavoči su najvjerojatnije rapidno migrirali te naselili istočni Tihi ocean, Atlantski ocean te Mediteransko more; skupina Gobiidae stigla je do Atlantskog preko zapadnog Tethys oceana, gdje je naknadno izumrla, dok se skupina Gobionellidae nije proširila iz oceana Tetisa dalje od SI Atlantika (Thacker, 2015).

14 UVOD

Slika 3. Geografska karta koja prikazuje globalne migracijske rute linija (eng. lineages) unutar porodice Gobiidae, s ancestralnim mjestom nastanka smještenim u Indopacifiku. Linija Gobius naselila je područja izvan Indopacifika (Mediteransko more, sjeveroistočni i jugoistočni Atlantski ocean, zapadni Indijski ocean) iako geološka i hidrološka karta globalnoga kontinentalnog područja u kenozoiku nije izgledala kao što izgleda danas (Indijski ocean i Mediteransko more bili su spojeni, kao i zapadni Atlantski i istočni Tihi oceani). Preuzeto iz Thacker, 2015.

1.3. Bioakustika, znanost o zvukovima u prirodi

Komunikacijski signal je bilo kakva promjena/signal jedinke odašiljača koja izaziva odgovor u ponašanju (tj. etološkom kontekstu) jedinke primatelja (Otovic i Partan, 2009). Većina životinja komunicira koristeći istovremeno nekoliko različitih signala (kemijskih, zvučnih, vizualnih itd.) pomoću različitih senzornih kanala, a taj je proces poznat pod nazivom „multimodalna“ ili „multisenzorna komunikacija“ (Otovic i Partan, 2009). Mnoge životinje, uključujući sisavce, ptice, žabe, ali i kukce, ribe i paukove, stvaraju zvukove (tj. akustičke signale) kako bi prenijele važne informacije ili usmjerile svoje ponašanje, a modulacijom svojih akustičkih parametara reguliraju sadržaj, tj. skup informacija koje taj zvuk prenosi (Simmons, 2003; Ladich, 2015; Suthers i sur., 2016; Ladich i Winkler, 2017). Zvukovi, kao sredstvo komunikacije, prenose ključne biološke informacije za jedinku, kao što su lokacija (hrane, partnera, predatora, objekta u okolišu), identitet (vrste, jedinke ili spola), socijalni status (dominantnost ili podređenost), motivacija (strah, bijes, spremnost za parenje) i dr. (Simmons, 2003). Ukratko, zvuk je fizički poremećaj u nekom mediju (vodi ili zraku) nastao kao rezultat pomicanja molekula zbog mehaničke aktivnosti (Simmons, 2003).

15 UVOD

Najjednostavnije objašnjeno, zvuk se može vizualizirati kao niz periodičkih i sinusoidnih promjena tlaka vode ili zraka tijekom vremena (Slika 4). Zbog fizičkog poremećaja uzorkovanog npr. glasnicama u grkljanu ili bubnjem, molekule u mediju se zbijaju (kondenzacija) ili šire (rarefrakcija) oko svojega ravnotežnog položaja (eng. equilibrium), šireći se u smjeru longitudinalnih valova (Beeman, 1998; Simmons, 2003). U blizini izvora zvuka molekule u zraku/vodi fizički se pomiču od svog ravnotežnog položaja u obliku gibanja čestica (eng. particle motion), što se naziva komunikacija u blizini polja, dok se one čestice (kondenzacije i rarefrakcije) koje se šire daleko od mehaničkog izvora zvuka šire kroz medij u obliku valova pritiska/tlaka (eng. pressure waves) te predstavljaju komunikaciju u daljini polja (na udaljenostima od izvora zvuka većim od 1 ̶ 3 valne duljine, komunikacija u daljini polja dominira u odnosu na komunikaciju u blizini polja; Simmons, 2003). Navedene razlike imaju važno biološko značenje budući da neke životinje (pogotovo kralješnjaci) komuniciraju na velikim udaljenostima (šišmiši, kitovi, i dr.), pri čemu se koriste valovi pritiska koji se dalje šire kroz medij, a mogu ih detektirati organi za sluh (uši), dok kod riba i žaba slušni organi detektiraju oba tipa širenja čestica (Simmons, 2003; Ladich, 2014; Ladich i Schulz- Mirbach, 2016; Ladich i Winkler, 2017). Neki su od najbitnijih parametara zvuka (tzv. akustičke varijable), važnih za životinjsku komunikaciju i prijenos informacija: frekvencija (u Hercima, Hz), amplituda (dB), faza, period (milisekunda), valna duljina (cm) i trajanje (sekunda). Odnosi među pojedinim parametrima imaju ključnu biološku važnost za životinje budući da još uvijek nije poznato npr. u kojem su odnosu frekvencija i period, tj. kako životinje percipiraju zvuk: u obliku broja ciklusa (kondenzacije i rarefrakcije) u određenom periodu (frekvencija) ili prema trajanju pojedinog ciklusa (period) (Simmons, 2003). Nadalje, ne reagiraju sve životinje na isti spektar frekvencija. Primjerice, ljudi registriraju zvukove frekvencije od 20 do 20.000 Hz (zvuk frekvencije ispod 20 Hz naziva se infrazvuk, a iznad 20.000 Hz ultrazvuk), slonovi registriraju zvukove ispod 20 Hz, dok sluh šišmiša i kitova registrira zvukove frekvencije i do 150.000 Hz koji se koristi kao sonar za eholokaciju (Simmons, 2003, Ladich i Winkler, 2017). Nadalje, valna duljina (λ), kao jedna od važnih prostornih parametara zvuka koja se odnosi na udaljenost u prostoru između dva susjedna brijega zvuka, razlikuje se za zvukove stvorene u zraku (gdje je brzina širenja zvuka zbog gustoće medija 343 m s-1) u odnosu na one stvorene u vodi (brzina zvuka iznosi 1.500 m s-1, uz minimalno prigušenje s udaljenosti), zbog čega je logično da će akvatičke životinje (kitovi i perajari) koristiti upravo zvukove manjih frekvencija (ali i dulje valne duljine) u odnosu na šišmiše (Simmons, 2003; Ladich i Winkler, 2017). Au i Hastings (2008) navode kako je upravo zvuk, zbog svoje sposobnosti dalekog raspršivanja i prelaska velikih udaljenosti, kao i

16 UVOD male apsorpcije, idealan tip komunikacijskog signala u vodenim ekosustavima. Navedene razlike u valnoj duljini imaju znatan utjecaj na sluh akvatičkih životinja i na lokalizaciju zvuka, ponajprije zbog odnosa valne duljine i razlike u udaljenosti slušnih organa (tj. ušiju), zbog čega neke životinje ne mogu registrirati zvukove velikih valnih duljina (Simmons, 2003).

Bioakustika (od grč. bios – život; akoustikos – slušno, čujno) je interdisciplinarna grana znanosti koja povezuje mnoge druge znanstvene discipline (zoologiju, animalnu fiziologiju, etologiju, akustiku – fiziku, i dr.), a proučava životinjske zvukove (uključujući i ljudske), tj. njihovo stvaranje, širenje (disperziju) i primanje (recepciju), utjecaj okoliša na akustičku komunikaciju, njegovu funkcionalnu povezanost s akustičkim parametrima životinjskih zvukova te, u novije vrijeme, utjecaj antropogene buke na fiziologiju i komunikaciju životinja (http://www-3.unipv.it/cibra/edu_underwaterbioacoustics_uk.html; Amorim i sur., 2018; De Jong i sur., 2018). Bioakustika proučava i fiziološke procese povezane sa stvaranjem i primanjem zvuka te istražuje kako funkcioniraju organi za primanje zvukova kod raznih životinja (većinom unutarnje uho, ali i kosti, čeljusti, otoliti itd. (Ladich i Schulz-Mirbach, 2016)). Razvoj bioakustike kao znanstvene discipline blisko je povezan s razvojem moderne tehnologije, pogotovo one za dokumentaciju zvukova (Beeman, 1998). Važna je prekretnica u tehnologiji (te ujedno i u bioakustici) postignuta oko 1950-ih godina, kada su razvijeni uređaji za snimanje, vizualizaciju i mjerenje akustičkih signala (mikrofon, hidrofon, osciloskop i spektrograf), dok je rapidni napredak u znanosti postignut razvojem digitalnih računala i naprednih matematičkih programa za modeliranje, prikaz i automatsku analizu zvukova (Beeman, 1998). Podvodna akustika je podgrana znanstvene discipline akustike koja proučava i istražuje širenje zvukova pod vodom, a razvila se isključivo zahvaljujući sonaru koji omogućuje detekciju objekata pod vodom i procjene dubine vode radi lakše navigacije (http://www-3.unipv.it/cibra/edu_underwaterbioacoustics_uk.html). Iako se većina tehnologije povezane s pasivnim sonarima i alatima za podvodno snimanje (npr. hidrofon) razvila zahvaljujući vojsci, u novije se vrijeme pasivni sonari koriste za detekciju i praćenje životinjskih zvukova, ponajprije onih koje stvaraju kitovi, ali i ribe, radi procjene njihove biomase te mjerenja globalne prosječne temperature i njenih fluktuacija u okolišu (http://www-3.unipv.it/cibra/edu_underwaterbioacoustics_uk.html; Urick, 1983; Au, 1993; Richardson i sur., 1995). U novije vrijeme razlikujemo dva tipa podvodne akustike: a.) aktivnu, koja koristi snop zvučnih valova stvorenih pomoću električnih pretvarača (eng. beamers) i primjenjuje akustičke metode raspršenja (eng. scattering methods) na organizmima

17 UVOD radi monitoringa i vizualizacije riba i ribljih populacija na određenim dubinama i b.) pasivnu, koja se temelji na hidrofonu te tehnikama snimanja i slušanja zvukova riba (i drugih vokalnih akvatičkih životinja) radi dobivanja detaljnih informacija o njihovoj distribuciji i ponašanju (Mann i sur., 2008). Kao osnovna pretpostavka pasivne akustike navodi se sposobnost ribe da producira zvuk te je kao takva podosta ograničena na one vrste koje se aktivno (ili pasivno) glasaju, ali i na lokaciju i vrijeme kada se one glasaju u svome okolišu i staništu (Mann i sur., 2008). Aktivna akustika znatno je revolucionalizirala komercijalni ribolov tako da je omogućila korištenje sonara za ehosondiranje (sustav u kojem jedna sonda generira zvukove, dok jedan prijamnik ili više njih „sluša“, tj. hvata jeku koja nastaje odbijanjem od površina) kako bi se lokalizirale razne vrste riba (Sund, 1935; Balls, 1948). Dvije su osnovne mjere koje primjena ehosondara omogućuje povratni radarski signal jedinki koje su sondirane (tj. njihova gustoća iz koje se izračunava i brojnost) te volumen povratnog radarskog signala (Mann i sur., 2008). Pasivna akustika istražuje zvukove riba i njihov etološki kontekst koji je većinom povezan s agresivnim interakcijama i reprodukcijom (udvaranje i parenje) radi dobivanja informacija kada se odvija mrijest pojedinih vrsta te koliko je jedinki prisutno na određenom području (Mann i sur., 2008). Dok je u aktivnoj podvodnoj akustici sonar najbitniji tehnološki dio opreme, u pasivnoj je akustici hidrofon već desetljećima dio standardne aparature za dokumentaciju zvukova i prostorno-temporalnog položaja vokalnih riba (Mann i sur., 2008). Dok su sonari postali široko primjenjiv dio opreme za procjenu brojnosti riba, pasivna akustika, tj. hidrofoni, to dosad još uvijek nije postigla. Razlog je vjerojatno taj što sustav za snimanje nije niti kompaktan niti komercijalno dostupan te što se podaci dobiveni pasivnom akustikom ne mogu primijeniti u „stručne svrhe“ (procjena brojnosti riba, količina jaja i fekunditet ženki itd.) koje bi koristili agencije i konzervacijske tvrtke te zbog velike količina danas neizbježne antropogene buke koja otežava posao bilježenja i interpretacije podataka (Mann i sur., 2008). Pasivna akustika, kao primjenjiva i relativno jeftina metoda, danas se koristi širom svijeta u laboratorijske svrhe, u sklopu kojih stručnjaci pokušavaju dokumentirati i opisati zvukove riba u kontroliranim (laboratorijskim) uvjetima, istražiti etološki kontekst produkcije zvukova i fiziološke odgovore jedinki na akustičku produkciju. Kao što je već navedeno, osnovni je dio opreme u eksperimentima pasivne akustike hidrofon, koji pretvara zvukove ribe u električni impuls (napon, tj. razliku električnog potencijala u milivoltima) koji se poslije može obrađivati, analizirati te slušati u stvarnom vremenu (Mann i sur., 2008). Metode obrade zvuka podrazumijevaju pretvaranje analognog signala u digitalni te generiranje triju najvažnijih prikaza za opisivanje zvuka: spektrograma/sonograma (promjena frekvencije zvuka i temporalnih karakteristika tijekom trajanja zvuk), oscilograma

18 UVOD

(promjena trajanja zvuka u odnosu na amplitudu) i spektra frekvencije (promjena frekvencije zvuka u odnosu na amplitudu) (Mann i sur., 2008; Fine i Parmentier, 2015). Istodobno, sustav za pohranu podataka (većinom audiosnimke .wav formata) bitan je dio opreme koji može izravno biti povezan s hidrofonom ili bežičnom vezom ili, alternativno, putem prijenosnog računala (laptop s internom memorijom) koji omogućuje praćenje zvukova u stvarnom vremenu (Mann i Lobel, 1995). Negativne su strane pasivne akustike velike količine generiranog, sirovog materijala, tj. audiosnimki na kojima se ne nalaze ciljani zvukovi te nemogućnost povezivanja zvuka s videosnimkom kako bi se utvrdila vrsta životinje koja je proizvela zvuk (ako se hidrofoni postavljaju na morsko dno bez vizualnog nadzora kao što je to u laboratorijskim eksperimentima), tako da bi ubuduće napredne metode trebale razviti automatizirani sustav kojim bi se omogućilo automatsko identificiranje vrsta (i jedinki) životinja koje stvaraju zvuk i mjerenje njihovih akustičkih parametara (Mann i sur., 2008; Fine i Parmentier, 2015). Budući da većina riba stvara tihe zvukove (maksimalna zabilježena amplituda zvukova sjenki iz porodice Sciaenidae iznosi 160 dB re: 1 Pa na 1 m udaljenosti) niske frekvencije (do 2 kHz), njihov spektar preklapa se sa spektrom frekvencija okolne, često antropogene buke (brodovi, seizmička istraživanja), što dodatno otežava njihovo bilježenje (Amorim, 2006; Mann i sur., 2008; Ladich, 2014; Fine i Parmentier, 2015).

Slika 4. Primjer isječka audiosnimke (.wav format) na kojem se vidi niz tonalnih zvukova proizvedenih od mužjaka glavočića okrugljaka (Neogobius melanostomus). A. niz zvukove se

19 UVOD prikazuje u obliku oscilograma (vrijeme u sekundama vs. amplituda u miliVoltima), dok se na B. oni prikazuju spektrogramom ili sonogramom (frekvencija u kHz vs. vrijeme u sekundama). Jačina intenziteta boje upućuje na količinu akustičke energije prisutne u zvuku (oko 90 Hz), koja je prisutna do otprilike 500 Hz. C. prikazuje detaljnu strukturu tonalnog zvuka na kojem se vidi kako je zvuk sastavljen od periodičkih sinusoidalnih valova. Autor: Horvatić, S.

Kao izvedena grana iz podvodne akustike razvila se i podvodna bioakustika, interdisciplinarna grana koja istražuje podvodne zvukove i akustičke parametre životinja i njihovo akustičko ponašanje radi praćenja (monitoringa) te utjecaj okolišnih parametara (klima, morske struje, temperatura itd.) na produkciju i širenje zvukova, ali i razvoj novih tehnologija i instrumenata za njihovo podvodno širenje i praćenje (http://www- 3.unipv.it/cibra/edu_underwaterbioacoustics_uk.html). Ubrzan razvoj i znanstvena popularnost podvodne bioakustike može se pripisati hidrofonu (podvodnom mikrofonu), sofisticiranoj napravi koja omogućuje snimanje, detekciju i pohranu životinjskih zvukova. Ukratko, hidrofon registrira nadolazeći zvučni val te ga pretvara u električni signal (zbog čega se može smatrati električnim pretvaračem), dok je većina današnjih hidrofona omnidirektnog smjera, odnosno mogu registrirati zvukove iz svih smjerova te registrirati širok spektar frekvencija (od nekoliko pa da 100 kHz) (http://www- 3.unipv.it/cibra/edu_underwaterbioacoustics_uk.html). Današnja je primjena hidrofona raznolika, a većinom se koriste u eksperimentima pasivne akustike ili u obliku dipolarnog transekata tijekom uzorkovanja morskih staništa, u kojima se nekoliko hidrofona spaja u dugačak niz (sustav prijamnika) razdvojenih svega nekoliko metara, čime se može odrediti smjer i vremenski dolazak (s vremenskim zakašnjenjem) zvuka svakom pojedinačnom hidrofonu. Ako se povežu s uređajem za pohranu podataka velike veličine (ili u novije vrijeme imaju ugrađenu memorijsku karticu u svom hardveru) i automatsku analizu zvukova, hidrofoni služe kao iznimno prigodno autonomno sredstvo namijenjeno za pasivnu svrhu (npr. mogu se spustiti na morsko dno te određeno vrijeme skupljati i pohranjivati podatke o vokalnim aktivnostima životinja (http://www- 3.unipv.it/cibra/edu_underwaterbioacoustics_uk.html; Luczkovich i sur., 2008; Mann i sur., 2008) (Slika 5).

20 UVOD

Slika 5. Primjena raznih metoda podvodne akustike (aktivna i pasivna akustika) za detekciju, dokumentaciju i monitoring životinjskih zvukova u Atlantskom oceanu te procjenu utjecaja antropogenih zvukova i buke na akvatičke životinje. Zvukovi se bilježe pasivnom metodom pomoću autonomnih snimača (montiraju se na morsko dno ili u vodeni stupac; A i B) ili se registriraju aktivno, pomoću akustičkih oznaka (eng. acoustic tags; C), podvodnom jedrilicom (eng. autonomous underwater glider; D) i sustavom hidrofona povezanim u niz (eng. array of hydrophones; E). Preuzeto i prilagođeno s https://www.fisheries.noaa.gov/new-england-mid-atlantic/endangered-species- conservation/passive-acoustic-research-atlantic-ocean.

1.4. Bioakustika riba

Bioakustika riba posebna je grana bioakustike (i podvodne akustike) koja istražuje produkciju i detekciju zvukova riba, funkcionalnu povezanost akustičke komunikacije i sluha (orijentacije) te evolucijske sile i selektivne pritiske koji utječu na razvoj navedenih struktura kod riba (Ladich, 2014). Iako je sasvim logično zaključiti kako su organi za stvaranje zvukova (sonički ili vokalni organi), koji su kod riba najraznolikiji u odnosu na sve kralješnjake, povezani s akustičkom komunikacijom, i dalje je nedovoljno poznato što to zapravo ribe slušaju osim zvukova srodnih i nesrodnih vrsta (Ladich, 2014, 2019; Parmentier, i Fine, 2016; Parmentier i sur., 2017; Fine i Parmentier, 2017). Također, rezultati mnogih istraživanja pokazuju kako su se sluh i komunikacija zvukovima kod riba razvili sasvim neovisno, na što upućuje velika raznolikost i broj različitih mehanizama za stvaranje zvukova (Ladich, 2015; Wysocki i Ladich, 2002). Taj se paradoks može objasniti sljedećim primjerom. Istražujući

21 UVOD vokalni repertoar ribe Trichopsis vittata (Cuvier, 1831), Henglmüller i Ladich (1999) i Wysocki i Ladich (2001) otkrili su kako mlade jedinke T. vittata (do 0,3 g) stvaraju već nakon 87. dana razvoja zvukove nazvane „prasak“ zvukovima (eng. bursts) pomoću posebno građenih tetiva prsnih peraja koje pri produkciji zvukova vibriraju kao žice na gitari. Kada su usporedili audiogram (slušnu krivulju koja upućuje na područja najistančanijeg sluha) sa spektralnim vrijednostima frekvencije produciranih zvukova (akustička energija prisutna na pojedinim frekvencijama) mladih i odraslih jedinki (Slika 6), autori su uvidjeli kako se dvije krivulje u mlađim uzrasnim kategorijama ne preklapaju (do 0,3 g), dok je to slučaj kod odraslih riba (od 0,6 g), što ih je navelo na zaključak kako se mlade jedinke međusobno ne čuju te ne mogu komunicirati zvukovima (Wysocki i Ladich, 2001). Iz toga su Wysocki i Ladich (2001) zaključili kako se ontogenetski kod T. vittata (a vjerojatno i svih ostalih riba) najprije razvija sluh, a naknadno sposobnost za produkciju akustičkih signala, što im tek poslije omogućuje intraspecijsku razmjenu informacija pomoću zvukova, tj. akustičku komunikaciju (Slika 6.).

Slika 6. Razvoj akustičke komunikacije kod zrakoperke Trichopsis vittata. A. Shematski prikaz ontogenetskog razvoja akustičke komunikacije kod zrakoperke T. vittata. Inicijalno, kod malih se jedinki T. vittata razvija sluh na određene frekvencije zvuka (B. – crna krivulja), nakon čega se poslije javlja sposobnost jedinki za produkciju akustičkih signala (B. – plava krivulja). Tek se u kasnijim fazama ontogenetskog razvoja audiogram najveće osjetljivosti zvuka (C. – crna krivulja) poklapa sa

22 UVOD spektrom najveće frekvencije (i amplitude) zvukova (C. – zelena krivulja). Preuzeto iz Ladich, 2015. Crne strelice na audiogramu B. i C. prikazuju područje najboljeg sluha, a plava/zelena strelica označuje područje najveće energije zvuka pri određenim frekvencijama (spektar snage, DF). DF – dominantna frekvencija (kHz)

Iako je kod šaranki malo vrsta vokalno, tj. samo nekoliko vrsta stvara zvukove, one imaju napredne strukture za poboljšavanje sluha (npr. Webberov uređaj), zbog čega se smatraju specijalistima. S druge strane, glavoči nemaju dodatne slušne strukture, a gotovo su sve istraživane vrste vokalne (Ladich, 2014). Sposobnost recepcije zvukova kod svih kralješnjaka temeljena je na osjetnim stanicama s osjetnim dlačicama u unutarnjem uhu iako su se dodatne slušne strukture namijenjene primanju zvukova te mehanizmi za njihovu produkciju najvjerojatnije razvili nezavisno kod danas opisanih riba (Ladich i Popper, 2004; Ladich, 2000, 2017). Kod riba nalazimo najraznolikije mehanizme za stvaranje zvukova (zbog čega se često nazivaju kukcima među kralješnjacima) te, za razliku od ostalih kopnenih kralješnjaka gdje se mehanizam temelji isključivo na vibraciji membrana (npr. glasnica u grkljanu ili pjevalu), kod riba ne postoji jedinstveni primarni organ za stvaranje zvukova (Ladich i Fine, 2006; Ladich, 2014; Parmentier i Fine, 2016). Međutim, akustička komunikacija istražena je samo kod riba zrakoperki, dok je za paklare, morske pse i dvodihalice podosta toga nepoznato, uz tek nekoliko istraživanja provedenih do danas, zbog čega se smatra kako je akustička komunikacija kod riba i tretrapodnih kralješnjaka rezultat konvergentne evolucije (Thompson, 1968; Ladich i Fine, 2006; Bass i Chagnaud, 2012; Christensen i sur., 2015). Jedan su od prvih popisa akustički aktivnih riba i njihovih zvukova objavili, za potrebe američke mornarice, Fish i Mowbray (1970), a rezultati su pokazali kako većina riba stvara specifične zvukove ne uzimajući u obzir ekološki kontekst (položaj ribe i interakcije s drugim jedinkama) pri produkciji zvukova. Uz to, samo je nekoliko istraživanja pokušalo rekonstruirati filogenetske odnose između vokalnih vrsta riba (Pomacentridae, Gobiidae, Batrachoididae) koristeći samo akustičke parametre u analizama (Malavasi i sur., 2008; Parmentier i sur., 2009; Rice i Bass, 2009), a rezultati tih istraživanja pokazali su kako neke vrste imaju specifične zvukove koji prenose filogenetski signal (tj. skup specifičnih informacija koji otkrivaju filogenetsku pripadnost pojedine vrste) zaslužan za njihovu diferencijaciju od ostalih, često genetski srodnih vrsta. Nasuprot tome, kod nekih skupina, kao što su ribe klaunovi (Amphiprioninae, Pomacentridae), zvukovi ne odvajaju srodne vrste, najvjerojatnije zbog selektivnog utjecaja veličine tijela i konteksta produkcije zvukova (samo

23 UVOD teritorijalni) na akustičke, ponajprije spektralne, parametre, zbog čega se maskira njihov filogenetski signal (Colleye i Parmentier, 2012). Ukratko, nekoliko je glavnih tipova mehanizama za stvaranje zvukova kod riba iako ih je, zbog svoje iznimne raznolikosti, teško striktno definirati (Slika 7): 1.) sonički („bubnjajući“) mišići povezani s plivaćim mjehurom (direktni – unutarnji mišići koji se povezuju sa stjenkom plivaćeg mjehura (Slika 7A); indirektni – vanjski mišići vezani za kost te povezani s plivaćim mjehurom; mišići koji dovode do vibracije mjehura bez funkcionalne veze s njegovom stijenkom, već su povezani tetivama ili koštanim pločama Slika 7B); 2.) prsne peraje i oplećje (stridulacijski zvukovi stvaraju se pri prelasku bodlje prsne peraje preko oplećja kao kod npr. somovki, a mišići imaju dodatnu funkciju u plivanju), Slika 7C; kod vrsta roda Trichopsis dvije tetive vibriraju tijekom pomicanja peraja, Slika 7D; kod peševa cijelo oplećje vibrira kontrakcijama posebnih mišića pri produkciji zvukova, Slika 7E); 3.) mehanizmi koji ne pripadaju nijednoj od tih dviju grupa (ždrijelni zubi (Slika 7E) i ostale strukture). Zvukovi riba stvoreni navedenim mehanizmima sastoje se od pulseva (ili niza pulseva), čija akustička energija ovisi o tipu mehanizma koji ih je generirao: zvukovi stvoreni plivaćim mjehurom imaju harmonijsku strukturu, a njihova akustička energija na sonogramu ne prelazi više od 500 Hz, s vrijednostima fundamentalne frekvencije (oko 200 Hz) jednake rapidnim kontrakcijama mišića plivaćeg mjehura; bubnjajući (eng. drumming) zvukovi koji traju od nekoliko milisekundi do nekoliko minuta kao kod vrste Portichthys notatus; stridulatorni zvukovi stvoreni prsnim mehanizmom (peraje i oplećje) čija akustička energija na sonogramu prelazi 1 kHz; niskofrekvencijski zvukovi (< 200 Hz) stvoreni ždrijelnim zubima ili vibracijama oplećja; tonalni zvukovi stvoreni rapidnim i kontinuiranim serijama mišićnih kontrakcija (fundamentalna frekvencija i stopa mišićnih kontrakcija se preklapaju), s vrijednostima energije oko 200 – 300 Hz (za sve reference, pogledati Ladich, 2014). Anatomski, svi kralješnjaci, uključujući i ribe, imaju organe za recepciju zvukova, ali nema svaka vrsta i organe za njihovo stvaranje. Morfološki, organi za sluh, tj. recepciju, mogu se podijeliti na isključivo unutarnje uho i dodatne periferne organe za primanje zvukova ili poboljšavanje sluha povezane s unutarnjim uhom (Ladich, 2014). Unutarnje uho, kao zajednički organ svih kralješnjaka, sastoji se od tri polukružna kanalića (osim kod sljepulja i paklara gdje se sastoji od dva kanalića) i tri vrećice, od kojih svaka posjeduje karbonatni otolit (slušni kamenčić) prislonjen na niz trepetljikinih stanica. Budući da su ribe jednake gustoće kao okolni medij, zvučni val prolazi kroz njih i one se kreću u istom smjeru kao nadolazeći val. Jedina teža struktura koja zaostaje za tim pokretima i dovodi do nagibanja trepetljika jest otolit (grč. ὠτο-, ōto- uho; λῐ́θος, líthos, kamen), zbog čega se stvara razlika u njihovu pomaku u odnosu na

24 UVOD zvučni val, što se poslije u mozgu registrira kao niskofrekvencijski zvuk (Hawkins, 1993; Popper, 2011; Ladich, 2014). Kod pojedinih riba postoji funkcionalna (direktna ili indirektna) veza unutarnjeg uha sa strukturom ispunjenom plinom (plivaći mjehur), što im omogućuje da detektiraju promjene u akustičkom tlaku (od nekoliko Hz do nekoliko kHz) koje se u obliku vibracija plivaćeg mjehura, kod nekih preko koštanih struktura (Webberov aparat), prenose do unutarnjeg uha (Fay, 1988; Ladich, 2004), dok neke ribe, kao pripadnici skupine Anabantoidei, imaju suprabranhialnu komoru (Hawkins, 1993). Odgovor na pitanje zašto su se mehanizmi za stvaranje zvukova razvili kod većine bentičkih, ali ne i kod svih pelagičkih skupina riba nalazi se najvjerojatnije u njihovim životnim navikama; da bi riba bila vokalna i stvarala zvukove, ona mora braniti limitirani resurs (npr. teritorij, hranu itd.) od potencijalnih napadača te privlačiti partnere (većinom ženke) u nastambu ili gnijezdo (pogotovo kod speleofilnih ili litofilnih vrsta), zbog čega im zvukovi služe kao sredstvo zastrašivanja u slučaju agresivnih interakcija ili omogućuju veći reproduktivni uspjeh, tj. fitnes za vrijeme reproduktivnih interakcija (Ladich i Myrberg, 2006; Ladich, 2014). Kod nekih manjih porodica riba (Doradidae, Bagridae, Pimelodidae, Batrachoididae, Gadidae, Sciaenidae, Holocentridae, Pomacentridae i Carapidae) sve su vrste, ili gotovo sve, vokalne, dok je kod velikih skupina, kao što su šaranke (Cyprinidae), većina vrsta nijema s nekoliko izuzetaka (Johnston i Johnson, 2000; Holt i Johnston, 2014; Fine i Parmentier, 2015).

25 UVOD

Slika 7. Raznolikost mehanizama za stvaranje zvukova kod riba i spektrogrami/sonogrami zvukova stvorenih tim mehanizmima. A. unutarnji mišići (SMi) koji se povezuju sa stijenkom plivaćeg mjehura (SL) kod Halobatrachus didactylus; B. vanjski mišići (SMe) koji su vezani na drugo rebro (2R) te na tetivu (BT) na ventralnoj strani plivaćeg mjehura kod piranje Serrasalmus rhombeus; C. stridulacijski mehanizam kod somića kod kojeg se dorzalni izdanak (DP) prsne bodlje (PS) tare po oplećju (PG); D. poboljšane tetive prsne peraje (ETs) koje vibriraju kao žice gitare kod Trichopsis vittata; E. mehanizam ždrijelnih zuba (PT) kod češljoustki, sunčanica i dr. te mehanizam temeljen na vibracijama oplećja pomoću soničkih mišića (SM) kod peševa. SB – plivaći mjehur, VC – kralješnica, BT – tetiva, PM – prsni mišić. Preuzeto iz Ladich (2014).

1.5. Pregled najvažnije literature o bioakustici glavoča

Glavoči (Gobiidae i Gobionellidae) su jedna od najistraživanijih skupina vokalnih riba (zajedno s Pomacentridae, Batrachoidiidae, Gadidiae, Serrasalmidae) iako je zanimljivo kako nemaju nikakve specijalizirane mehanizme ili strukture za akustičku komunikaciju i primanje zvukova (npr. neke vrste nemaju čak ni plivaći mjehur (Ladich, 2014; Zeyl i sur., 2016)). Plivaći mjehur predložen je kao osnovni element mehanizma za produkciju zvukova kod mnogih skupina riba (Ladich i Fine, 2006), dok se kod glavoča pokazalo kako vrste koje imaju (Padogobius bonelli, Pomatoschistus minutus (Pallas, 1770), Gobius cruentatus Gmelin, 1789) ili nemaju (Padogobius nigricans, Neogobius melanostomus) navedenu strukturu i dalje imaju sposobnost vokalizacije (Zeyl i sur., 2016). Do danas su 24 vrste (oko 1,2% ukupne brojnosti, Fricke i sur., 2020) istražene u bioakustičkim eksperimentima, sa samo jednom vrstom (Economidichthys pygmaeus (Holly, 1929)) koja se nije glasala tijekom istraživanja (Gkenas i sur., 2010; Zeyl i sur., 2016). Zajedno s drugim vokalnim skupinama riba (Cottidae i Percidae) glavoči dijele konvergentne životne navike (bentički način života i speleofilni tip reprodukcije), zbog čega su jako dobri modelni organizmi u eksperimentima pasivne akustike, evolucijske biologije ili akustike općenito (Zeyl i sur., 2016). Oni stvaraju niskofrekvencijske zvukove (oko 200 – 300 Hz) tijekom agresivnih ili reproduktivnih interakcija, dok su od posebne važnosti akustički signali stvoreni tijekom parenja, za koje se smatra kako povećavaju reproduktivni uspjeh (tj. fitnes) mužjaka zbog utjecaja seksualne selekcije na akustička svojstva (Zeyl i sur., 2016). Eksperimenti povezani s anatomskom strukturom glavoča pokazali su kako plivaći mjehur najvjerojatnije nije uključen u proces produkcije zvukova (Lugli i sur., 2003; Parmentier i sur., 2013), a zbog morfoloških sličnosti kranio-pektoralnih mišića (onih vezanih za lubanju i prsne peraje) glavoča i peševa (Parmentier i sur., 2013; Colleye i sur., 2013), zajedno s dokumentiranim pokretima glave

26 UVOD

(klimanje) i prsnih peraja tijekom produkcije zvukova, smatra se kako se mehanizam za stvaranje zvukova temelji na kontrakcijama navedenih mišića. Naime, kako bi poboljšali transmisiju niskofrekvencijskih zvukova kroz akustički ograničeni okoliš (plitka voda ponaša se kao filtar za zvukove niskih frekvencija (Rogers i Cox, 1988; Mann, 2006)), glavoči iskorištavaju dva ekološka faktora: njihovi niskofrekvencijski zvukovi gotovo savršeno odgovaraju „tihom prozoru“ u frekvencijskom spektru okolišne buke (oko 150 – 300 Hz, Lugli i sur., 2003; Lugli, 2010) te amplificiraju frekvencije zvukova nastambama u kojima se razmnožavaju (Lugli, 2012, 2013, 2014). Zvukovi glavoča mogu se podijeliti s obzirom na vrijeme kada ih stvaraju, npr. tijekom agresivnih interakcija (mužjak – mužjak ili mužjak – ženka), za vrijeme privlačenja ženke koja je izvan nastambe (eng. courtship) i tijekom parenja u nastambi (eng. pre-spawning i spawning Lugli i sur., 1997; Lugli i Torricelli, 1999; Johnston i Johnson, 2000; Myrberg i Lugli, 2006). Prethodno spomenuti zvukovi ne šire se daleko u plitkim (obalnim) vodama, zbog čega se ne koriste za dalekometnu komunikaciju (Zeyl i sur., 2016). Od ukupnog broja istraživanih glavoča, njih 16 stvara isključivo pulsatilne zvukove u agresivnim ili reproduktivnim interakcijama, dok je kod ostalih vrsta akustička struktura zvukova tonalnog karaktera (udvaranje) ili kao dio kompleksnog zvuka (Lugli i sur., 1997). Kod malog broja vrsta zvukovi znaju biti popraćeni tupim zvukovima (eng. thump sounds) stvorenim tijekom udvaranja izvan nastambe kada mužjaci prilaze ženki (Amorim i Neves, 2007; Malavasi i sur., 2009). Većina vokalnih glavoča pripada porodicama Gobiidae i Gobionellidae (filogenija prema Thacker, 2009; Thacker i Roje, 2011), dok je jedina vokalna vrsta izvan navedenih skupina Odontobutis obscura (Temminck i Schlegel, 1845) (Odontobutidae) za koju su dokumentirani isključivo pulsatilni zvukovi (Takemura, 1984). Takvi rezultati, kao i veća učestalost pulsatilnih zvukova u odnosu na tonalne, naveli su neke autore na zaključak kako se sposobnost za produkciju zvukova kod glavoča u širem smislu (podred Gobioidei) javlja filogenetski dosta rano i kako su tonalni zvukovi odvedeno svojstvo budući da je porodica Odontobutidae sestrinska skupina svim drugim porodicama podreda Gobioidei (uključujući Gobiidae i Gobionellidae; Thacker, 2009; Agorreta i sur., 2013; Zeyl i sur., 2016). Slijedeći navedenu akustičku hipotezu, tonalni zvukovi mogli su potencijalno nastati od pulsatilnih tako da je u zvuku nestao period između dva pulsa, zbog čega se oni ponavljaju velikom brzinom, a zvukovi kratko traju te se moduliraju u frekvenciji (Malavasi i sur., 2008).

Jedno su od prvih detaljnih istraživanja vokalnog repertoara glavoča mediteranskog područja proveli Torricelli i sur. (1990). Autori su snimili zvukove te kvantitativnom i

27 UVOD kvalitativnom analizom opisali strukturu akustičkih signala teritorijalnih mužjaka slatkovodnog glavočića (Padogobius bonelli) te dodatno istražili kakav utjecaj okolišni (temperatura vode) i individualni faktori imaju na zvukove i njihova svojstva, pod hipotezom da kod ektotermnih organizama imaju znatan utjecaj na fiziologiju. Padogobius bonelli slatkovodna je vrsta autohtona za sjevernu Italiju (porječje rijeke Po), Švicarsku, Sloveniju i Hrvatsku, gdje naseljava male potoke s kamenim/šljunkovitim supstratom, a razmnožava se u kasno proljeće kada ženka polaže jaja na krov nastambe, a mužjak pokazuje paternalnu brigu za mlade (Miller, 2004). Zvukovi mužjaka P. bonelli bili su dokumentirani tijekom agresivnih (mužjak – mužjak) i reproduktivnih (udvaranje) interakcija, a kvalitativna je analiza pokazala kako se oni sastoje od rapidno produciranih pulseva, sa stopom ponavljanja koja se smanjuje kako zvuk odmiče kraju (Torricelli i sur., 1990). Nadalje, rezultati istraživanja pokazali su kako temperatura vode ima znatan utjecaj na pojedine akustičke parametre zvuka (stopu ponavljanja pulseva, modulaciju i trajanje) s posebnim naglaskom na stopu ponavljanja pulseva (prosječno oko 300 Hz) koja je negativno korelirana, dok je trajanje pozitivno korelirano s temperaturom. Veličina tijela (samim time i veličina mišića za stvaranje zvukova) nije bila znatno povezana s akustičkim parametrima, dok su agresivni zvukovi trajali dulje i imali manje vrijednosti stope ponavljanja pulseva (i fundamentalne frekvencije) od reproduktivnih zvukova, koji su se pokazali manje varijabilnima od agresivnih signala (Torricelli i sur., 1990). Autori zaključuju kako su male varijabilnosti reproduktivnih zvukova najvjerojatnije posljedica velike stereotipnosti signala uključenih u proces reprodukcije (Torricelli i sur., 1986; Torricelli i sur., 1990).

Lugli i sur. (1996) proveli su prvo komparativno akustičko istraživanje radi dokumentiranja i opisa akustičkih svojstava i zvukova triju talijanskih slatkovodnih glavoča (Padogobius bonelli, P. nigricans i O. punctatissimus) produciranih tijekom reproduktivnih intraspecijskih interakcija (udvaranje i parenje). Istovremeno, u sklopu laboratorijskih i in situ eksperimenata, autori su kvantitativno istražili etološki kontekst produkcije zvukova navedenih glavoča. Poznato je kako glavoči komuniciraju intraspecijski (mužjak – mužjak ili mužjak – ženka) koristeći istovremeno različite senzorne signale (vizualne, kemijske i akustičke, Tavolga, 1954), a kao osnovna pretpostavka provedenog rada bila je činjenica kako zvuk niskih frekvencija proizveden na malim udaljenostima potencijalno ima važnu ulogu u reprodukciji glavoča. Provedeno je istraživanje pokazalo kako je vokalni repertoar glavoča iznimno velik, a u slučaju slatkovodnog glavočića P. bonelli sastoji se od čak tri, akustički različita, zvuka (tonalni, pulsatilni i kompleksni) stvorenih tijekom različitih faza

28 UVOD reprodukcije. Tonalni zvukovi P. bonelli, zabilježeni ovim i prijašnjim istraživanjima (Torricelli i sur., 1986; Torricelli i sur., 1990), stereotipni su signali kratkog trajanja (oko 200 – 300 ms), čiji zbijeni pulsevi odaju izgled sinusoidnog vala produciranog zvuka. Nadalje, pulsatilni („bubnjajući“) zvukovi sastoje se od niza pulseva stvorenih u kratkom razdoblju (40 pulseva/sec.) koji traju između 50 i 800 ms te imaju maksimalne vrijednosti frekvencije oko 100 Hz. Kombinacijom navedenih zvukova, P. bonelli stvarao je kompleksne (dvokomponentne) zvukove koji se sastoje od početnog (pulsatilni, trajanje: 60 – 800 ms) i krajnjeg (tonalni, trajanje: 70 – 700 ms) dijela, kod kojega je većina akustičke energije koncentrirana na fundamentalnoj frekvenciji (Lugli i sur., 1996). Za razliku od P. bonelli, druge dvije vrste producirale su samo jedan tip zvukova: K. puncatitssima stvara samo pulsatilni tip (strukturalno sličan pulsatilnim zvukovima P. bonelli), karakteriziran s 40 pulseva/s, čije trajanje varira između 400 i 1000 ms; P. nigricans emitirao je samo kratke, stereotipne tonalne zvukove sinusoidalnog izgleda u trajanju 300 – 400 ms (Lugli i sur., 1996). S obzirom na etološki kontekst, tj. prisutnost ženke u blizini ili u gnijezdu, autori su signale podijelili na: 1. zvukove stvorene kada je ženka izvan gnijezda (eng. pre-spawning) i 2. zvukove stvorene kada je ženka u gnijezdu (eng. spawning, Lugli i sur., 1995, 1996). Zvukove prve grupe iz nastambe stvarali su mužjaci P. nigricans i P. bonelli, s naglaskom kako obje vrste stvaraju samo tonalne zvukove pri udvaranju maksimalnom stopom 60 – 90 zvukova/min. (Lugli i sur., 1996). Kada je ženka unutar gnijezda, mužjaci P. bonelli stvarali su pulsatilne i kompleksne zvukove tijekom cijelog njezina boravka, a prestaju tek kada ženka napusti gnijezdo. Budući da su se zvukovi stvarali tijekom cijelog razdoblja ženkina boravka u nastambi (a ne neposredno prije parenja), autori zaključuju kako oni nemaju funkciju u sinkronizaciji ispuštanja gameta (Lugli i sur., 1996). Kao i kod P. bonelli, mužjaci O. punctatissiumus stvarali su pulsatilne zvukove samo kada je ženka ušla u nastambu i trajali su sve dok je ona tamo boravila, sa stopom 2 – 10 zvukova/min u trenutku kada ženka zauzme ovipoziciju (pozicija za polaganje jaja, trbuhom okrenuta prema stropu nastambe). Kod obje vrste koje produciraju zvukove kada je ženka u nastambi (P. bonelli i O. punctatissiumus), najveći je broj zvukova mužjaka stvoren tijekom prvih 10 – 15 min nakon ovipozicije, a kako ženka polaže sve više jaja na strop nastambe, tako se produkcija zvukova smanjuje (Lugli i sur., 1996). Mužjaci P. bonelli nisu producirali zvukove jednom kada je ženka ušla ili boravila u gnijezdu, već su ih stvarali samo kada je ženka na ulazu ili kruži oko njega (Lugli i sur., 1996). Rezultati istraživanja upućivali su na to kako zvukovi kod glavoča ne sudjeluju u procesu sinkroniziranog izbacivanja gameta tijekom reprodukcije, već imaju „udvaračku“

29 UVOD funkciju budući da se ne stvaraju samo netom prije parenja ili fertilizacije jaja, već ih neke vrste produciraju prije ovipozicije i tijekom reprodukcije (Lugli i sur., 1996).

Prvo detaljno multidisciplinarno istraživanje mehanizma za produkciju zvukova glavoča proveli su Parmentier i sur. (2013). Cilj je istraživanja bio, koristeći morfološko- eksperimentalne analize i rezultate nekoliko različitih metoda (elektromiografija, anatomija, superbrze kamere, histologija, bioakustika), ispitati koje strukture sudjeluju u procesu produkcije zvukova te koji mehanizam najbolje opisuje akustička svojstva kod morske vrste glavoča Gobius paganellus, koji stvara dva tipa zvukova (pulsatilni i tonalni; Malavasi i sur., 2008). Gobius paganellus velika je vrsta morskog glavoča koja je rasprostranjena od Škotske pa sve do Senegala i Crnog mora (Miller, 1986), a u provedenom pokusu autori su akustički istražili i snimili dvije populacije (Francuska i Italija) u laboratorijskim uvjetima koristeći metodu pasivne akustike te kvalitativnom i kvantitativnom bioakustičkom analizom istražili akustička svojstva zvukova. Rezultati su pokazali kako se populacije znatno razlikuju s obzirom na tip produciranog zvuka: u Francuskoj G. paganellus je stvarao samo pulsatilne zvukove (trajanje: 192 ms; 11 pulseva; trajanje pulsa: oko 20 ms; puls period: 60 Hz) popraćene pokretima bukalne, brahialne i operkularne regije, dok to nije bio slučaj s talijanskim jedinkama koje su stvarale samo tonalne zvukove (trajanje: 334 ms; 31 puls; puls period: 93 Hz) bez uočenih pokreta glave i ostalih regija, zbog čega se uloga pokreta glave odbacila kao potencijalni mehanizam. Razlike u tipu produciranog zvuka (Francuska – pulsatilni naspram Italija – tonalni) pripisala se razlici u dužini tijela između dviju populacija (Parmentier i sur., 2013). Rezultati superbrzih kamera (200 okvira/s) pokazali su kako su ribe iz Francuske stvarale brze pokrete glavom tijekom produkcije zvukova (faza pripreme i faza spuštanja glave ili „klimanje“ glavom), čiji ciklusi savršeno odgovaraju periodu svakog pulsa njihovih zvukova, dok navedeni pokreti nisu bili primijećeni kod talijanske populacije, najvjerojatnije zbog razlike u tipu zvuka između dviju populacija (Parmentier i sur., 2013). Anatomska je analiza pokazala kako se lubanja i okolne strukture G. paganellus ne razlikuju između dviju populacija te kako je ona tipična za ribe reda Perciformes (Vandewalle i sur., 1982; Barel, 1983; Liem, 1993). Upotrebom elektromiografije i elektronske mikroskopije, pokazalo se kako mišić levator pectoralis, koji se veže na lubanju i povezuje na dorzalni dio oplećja (kosti grlenjače), sudjeluje u procesu stvaranja zvukova kod G. paganellus. Autori su, koristeći anatomske analize, navedeni mišić podijelili na dva različita snopa koja se križaju na mjestu vezanja na grlenjači: pars lateralis (drži se kaudalnog dijela neurokraniuma, a veže se na rostralni, tj. ventralni dio grlenjače) i pars medialis (drži se ventralnog dijela

30 UVOD neurokraniuma, a veže se dijagonalno na dorzalnoj strani grlenjače). S obzirom na dobivene rezultate anatomskih i bioakustičkih analiza, autori su predložili kako kontrakcijama navedenih mišića i vibracijama oplećja (ossi radiales) najvjerojatnije dolazi do vokalizacija kod jedinki G. paganellus (Parmentier i sur., 2013). Nadalje, histološka je analiza pokazala kako se poprečni presjek mišića levator pectoralis sastoji od tri zone bogate mitohondrijima i posebnog rasporeda miofibrila, dok je navedeni mišić tipične strukture i izgleda kao i ostali sonički mišići drugih vokalnih riba (najbrže kontrahirajući mišići među kralješnjacima; Loesser i sur., 1997; Ladich, 2001; Boyle i sur., 2013). Rezultati elektromiografije (EMG), tj. upotrebe para elektroda utaknutih u lijevi i desni mišić levator pectoralis kako bi se stimulirala njihova aktivnost, potvrdili su kako stopa kontrakcija mišića savršeno odgovara vrijednostima fundamentalne frekvencije te najvjerojatnije i određuje spektralne vrijednosti akustičkih signala (Parmentier i sur., 2013). Stoga su autori zaključili kako glavoči i peševi (Cottidae), zbog konvergentnih anatomskih, kinetičkih (pokreti glave, tj. klimanje glavom) i ekoloških svojstava (bentički način života i speleofilna taktika) dijele i jednaki mehanizam za stvaranje zvukova temeljen na kontrakcijama mišića levator pectoralis i okolnih struktura oplećja, odbacujući pretpostavku kako bi plivaći mjehur mogao imati ulogu u procesu glasanja (Parmentier i sur., 2013).

Jedno su od najvažnijih, a ujedno i prvo detaljnije interspecijsko istraživanje produkcije zvukova glavoča, proveli Malavasi i sur. (2008), u sklopu kojega su autori analizirali šest kvalitativnih i kvantitativnih akustičkih parametara zvukova (stvorenih u agresivnom i reproduktivnom kontekstu) za jedanaest vrsta mediteranskih glavoča rodova linije Gobius (Padogobius, Zosterisessor, Gobius; prema Agorreta i sur., 2013) i linije Pomatoschistus (Pomatoschistus i Knipowitschia; prema Agorreta i sur., 2013) te pokušali odrediti koliko se rezultati provedenih bioakustičkih analiza poklapaju s prijašnjim morfološkim i molekularnim istraživanjima (Miller, 1990; McKay i Miller, 1991, 1997; Miller i sur., 1994; Penzo i sur., 1998; Simonovic, 1999; Sorice i Caputo, 1999; Huyse i sur., 2004). Osnovna je hipoteza istraživanja bila da, zbog svoje velike raznolikosti i širokog vokalnog repertoara, glavoči mogu poslužiti kao idealni modelni organizmi u filogenetskim i bioakustičkim analizama koje istražuju utjecaj srodnosti ili adaptacija na evoluciju akustičke komunikacije. Rezultati su pokazali kako se, na temelju akustičkih parametara, sve vrste znatno razlikuju jedna od druge, što pokazuje kako su zvukovi specifični za vrstu (Malavasi i sur., 2008). Detaljna akustička i multivarijatna statistička analiza pokazale su kako su dva najvažnija akustička parametra zaslužna za razlikovanje vrsta upravo temporalna svojstva: trajanje zvuka (DUR) i stopa

31 UVOD ponavljanja pulseva (PRR) (Malavasi i sur., 2008). Takvi rezultati poklapali su se s prijašnjim istraživanjima nekoliko vokalnih porodica riba Pomacentridae, Centrarchidae, i Cichlidae (Lobel, 1998; Amorim i sur., 2004; Myrberg i Lugli, 2006) koja su potvrdila sličnu ulogu temporalnih svojstava u diferencijaciji vrsta. Uz to, tri su se grupe vrsta mogle razlikovati s obzirom na dva najvažnija akustička parametra: tonalna grupa (visok PRR, kratko DUR; Padogobius spp. i Gobius paganellus), velika pulsatilna grupa (nizak PRR, kratko DUR; Gobius spp. i Zosterisessor ophiocephalus) i mala pulsatilna grupa (nizak PRR, dugačko DUR; Pomatoschistus spp. i Knipowitschia spp.). Na fenogramu konstruiranom na temelju akustičkih svojstava, glavoči pijeska (Pomatoschistus, Orsinigobius i Knipowitschia) odvojili su se u jednu, dok su se ostale velike vrste odvojile u drugu grupu, unutar koje su rodovi Gobius i Padogobius prikazani parafiletski budući da se Padogobius nigricans grupirao s G. paganellus i G. cobitis, a P. bonelli s G. niger (Malavasi i sur., 2008). Zanimljivo, rezultati istraživanja pokazali su kako su vrijednosti modulacije frekvencije (FM) kod većine vrsta bile blizu nule ili negativne, dok je jedino kod G. paganellus FM iznosila oko 25 Hz, što je sugeriralo kako su zvukovi navedene vrste znatno povišeni u frekvenciji (Malavasi i sur., 2008). Nadalje, analiza korelacije između akustičkih parametara (DUR i PRR) i veličine tijela te temperature pokazala je kako je DUR znatno (negativno) korelirana s veličinom tijela i PRR kod istraživanih vokalnih glavoča. Autori su naglasili kako negativna korelacija DUR (koja odvaja dvije grupe glavoča na fenogramu) i veličine tijela te njihova predvidljivost (male vrste – dugački zvukovi i obrnuto) može imati filogenetsko značenje zbog morfološkog ograničenja koje predstavlja za mehanizam produkcije zvukova; evolucijske (selektivne) sile povezane s razlikama u veličini tijela glavoča mogle bi nastati kao rezultat razlika u trajanju zvuka koje je potencijalno svojstvo za razlikovanje identiteta riba (Kihslinger i Klimley, 2002; Malavasi i sur., 2008). Istovremeno, rezultati bioakustičkog istraživanja poklapali su se djelomično s morfološkim i molekularnim analizama, što je autorima sugeriralo kako akustički signali (zajedno sa svojim kvantitativnim parametrima) potencijalno prenose filogenetski signal zaslužan za razlikovanje vrsta (Malavasi i sur., 2008). Kako bi dobivene rezultate autori objasnili u filogenetskom kontekstu, autori su razvili hipotezu u obliku grafičkog prikaza, o evoluciji akustičke komunikacije kod glavoča (Slika 8): pulsatilni zvuk najvjerojatnije predstavlja ancestralan tip (pleziomorfno svojstvo) koji se zadržao kod svih vokalnih glavoča pijeska, kao i kod nekih velikih vrsta (Gobius, Zosterisessor, Padogobius), dok su tonalni zvukovi kod G. paganellus i Padogobius spp., u parsimonijskom scenariju, odvedeni tip zvukova nastali od starijih pulsatilnih tako što se njihovo trajanje (DUR) smanjilo, omogućivši povećanje stope ponavljanja pulseva (PRR, zvukovi se u kratkom

32 UVOD vremenu mnogo puta ponove, bez međuintervala) s obzirom na zapaženu negativnu međusobnu korelaciju DUR i PRR. Također, kako su zvukovi postali kraći s velikom stopom ponavljanja pulseva, tako su dobili mogućnost biti frekvencijski modulirani, što je naknadno otvorilo put za divergenciju bliskih vrsta (npr. G. paganellus od G. cobitis (Malavasi i sur., 2008)).

Slika 8. Dijagram koji prikazuje evoluciju akustičke komunikacije i razvoj zvukova kod mediteranskih glavoča (rodovi Padogobius, Zosterisessor, Gobius, Knipowitschia, Orsinigobius i Pomatoschistus). Prema navedenoj hipotezi, inicijalni je tip zvuka bio pulsatilni, koji je modulacijom dvaju najvažnijih akustičkih svojstva (DUR – trajanje; PRR – stopa ponavljanja pulseva) doveo do razvoja pulsatilnih/bubnjajućih zvukova kod skupine Pomatoschistus + Orsinigobius + Knipowitschia, tonalnih zvukova kod Padogobius spp. + G. paganellus te pulsatilnih zvukova kod preostale skupine. Na dijagramu, na mjestu gdje se povećava (>) PRR, javlja se ujedno i tonalna struktura zvukova. Preuzeto i prilagođeno iz Malavasi i sur. (2008)

33 PREGLED DISERTACIJE

PREGLED DISERTACIJE

Ova disertacija obuhvaća četiri izvorne znanstvene publikacije (I – IV) u kojima su navedeni ciljevi i znanstvene hipoteze.

Ciljevi ove disertacije:

1. Zabilježiti zvukove i kvantificirati temporalne i spektralne parametre akustičkih signala pontokaspijskih glavoča (Gobiidae) Neogobius fluviatilis Pallas, 1918, N. melanostomus (Pallas 1814) i Ponticola kessleri (Günther 1861), kao i rotana Perccottus glenii Dybowski 1877 (Odontobutidae) te detaljno opisati intraspecijsku akustičku varijabilnost zvukova istraživanih vrsta.

2. Istražiti interspecijsku akustičku varijabilnost vokalnih vrsta linije Gobius, tj. atlantsko- mediteranskih glavoča (rodovi Gobius, Padogobius i Zosterisessor) i pontokaspijskih vrsta (Neogobius fluviatilis, N. melanostomus i Ponticola kessleri), rekonstruirati hipotezu o evoluciji akustičke komunikacije glavoča linije Gobius te istaknuti najvažnija akustička svojstva koja diferenciraju vokalne vrste.

3. Istražiti etološki kontekst akustičke komunikacije i anatomske strukture povezane s potencijalnim mehanizmom za produkciju zvukova kod rotana P. glenii (Odontobutidae) te usporediti morfološke rezultate s predloženim mehanizmom kod ostalih ispitanih vrsta.

4. Istražiti genetsku srodnost vokalnih glavoča linije Gobius, rekonstruiranu na temelju DNK sekvenci četiri mitohondrijska i jezgrena biljega (cyt b, COI, Rho i RAG), te usporediti dobivene rezultate s provedenim akustičkim analizama kako bi se istražilo postoji li poklapanje akustičke divergencije s genetičkom.

Znanstvene hipoteze ove disertacije:

i. Filogenetska interspecijska analiza pontokaspijskih i atlantsko-mediteranskih glavoča iz linije Gobius, temeljena na kvalitativnim i kvantitativnim akustičkim parametrima, upućuje na određen stupanj međusobne srodnosti pojedinih vokalnih vrsta. Također, pretpostavlja se kako se vokalni repertoar vokalnih glavoča linije Gobius sastoji od malog broja različitih tipova zvukova. ii. Kod glavoča linije Gobius glasaju se samo mužjaci tijekom intraspecijskih interakcija. iii. Mehanizam za produkciju zvukova temelji se kod rotana Perccottus glenii (Odontobutidae) na kranio-pektoralnim mišićima, a samo pulsatilni zvukovi predstavljaju ancestralni tip akustičkog signala kod glavoča u širem smislu (Gobioidei).

34 PREGLED DISERTACIJE iv. Interspecijski filogenetski odnosi devet vokalnih vrsta glavoča linije Gobius (rodovi Neogobius, Ponticola, Zosterisessor, Gobius, Padogobius), konstruirani na temelju šest kvantitativnih akustičkih parametara, preklapaju se s rezultatima molekularnih analiza, što pokazuje kako zvukovi prenose filogenetski signal odgovoran za razlikovanje srodnih vrsta.

Publikacije I, II i III blisko su povezane s prvim ciljem te odgovaraju na prvu i treću hipotezu ove disertacije. Publikacija I opisuje zvukove i kvantificira akustičke parametre vrste Neogobius fluviatilis, dotad jedne od tek nekoliko ispitanih vrsta glavoča s pontokaspijskog područja. Rezultati navedenog rada pokazuju kako su tonalni zvukovi zajednička karakteristika ispitanih pontokaspijskih vrsta te kako nisu samo mužjaci ti koji stvaraju komunikacijske zvukove, već i ženke koje imaju sposobnost produciranja akustičkih signala. Publikacija II dodatno produbljuje znanje o akustičkoj komunikaciji pontokaspijskih vrsta Neogobius melanostomus i Ponticola kessleri te daje dublji uvid u akustičku divergenciju vrsta pontokaspijskog područja koje su prijašnjim morfološkim istraživanjima bile svrstane u jedinstveni rod Neogobius. Istovremeno, rezultati publikacije II pokazuju kako su tonalni zvukovi pontokaspijskih glavoča izrazito stereotipni i specifični za ispitane dvije vrste, s određenim parametrima akustičkih signala koji bi potencijalno mogli biti zaslužni za njihovu divergenciju. Nekoliko je različitih skupina prepoznato akustičkim analizama, na temelju akustičkih parametara, u sklopu publikacije II, a dvije srodne vrste (N. melansotomus i N. fluviatilis) dijele određene zajedničke karakteristike koje ih odvajaju od dotad dokumentiranih vrsta. Publikacija III izravno je povezana s trećim ciljem te pripada drugoj hipotezi ove disertacije. Naime, publikacija III daje dublji uvid u akustičku komunikaciju rotana P. glenii, člana bazalne porodice Odontobutidae unutar podreda Gobioidei. Rezultati te publikacije istaknuli su šest etoloških kategorija tijekom kojih dolazi do produkcije zvukova kod rotana za vrijeme reprodukcije te utvrdili kako nisu samo pulsatilni zvukovi sinapomorfno svojstvo za glavoče u širem smislu (Gobioidei), već da, uz pulsatilne, i tonalni zvukovi predstavljaju ancestralne akustičke signale. Uz to, publikacija III navodi anatomske strukture, tj. kranio- pektoralni mišić oplećja i lubanje podijeljen u tri snopa (m. levator pectoralis: pars lateralis superficialis, pars lateralis profundus i pars medialis) kao potencijalni mehanizam za produkciju zvukova kod rotana. Na kraju, publikacija IV izravno odgovara na prvu i treću hipotezu, a povezana je s drugim i četvrtim ciljem. Publikacija IV objedinjuje i analizira sve podatke (zvukove s njihovim akustičkim parametrima) dugogodišnjih akustičkih istraživanja, kao i ove disertacije, radi ispitivanja korelacije između akustičke i genetičke divergencije kod

35 PREGLED DISERTACIJE devet atlantsko-mediteranskih i pontokaspijskih vokalnih glavoča. Rezultati publikacije IV pokazuju kako su zvukovi kod vokalnih glavoča linije Gobius šireg mediteranskog područja izrazito specifični za svaku vrstu budući da su akustička divergencija i filogenetska diferencijacija pozitivno korelirane kod istraživanih vokalnih vrsta. Također, rezultati akustičkih analiza publikacije IV pokazuju kako određeni parametri zvuka najvjerojatnije prenose filogenetski signal zaslužan za razdvajanje srodnih vrsta. Filogenetski odnosi između vokalnih vrsta linije Gobius rekonstruirani su na temelju četiri mitohondrijska (mtDNA) i jezgrena (nDNA) genetička biljega, a rezultati navode kako je rod Padogobius parafiletskog podrijetla zbog bliske (utvrđene) genetske srodnosti talijanskog P. nigricans s pontokaspijskim vrstama Neogobius roda, kao i sličnog tipa akustičkog signala koje ove vrste produciraju.

36 PREGLED DISERTACIJE

37 ZNANSTVENE PUBLIKACIJE

2. ZNANSTVENE PUBLIKACIJE

XI ZNANSTVENE PUBLIKACIJE Biological Journal of the Linnean Society, 2016, 117, 564–573. With 3 ﬁgures.

Sound production in the Ponto-Caspian goby Neogobius ﬂuviatilis and acoustic afﬁnities within the Gobius lineage: implications for phylogeny

1 2 1 SVEN HORVATIC , FRANCESCO CAVRARO , DAVOR ZANELLA and STEFANO MALAVASI2*

1Department of Zoology, Faculty of Science, University of Zagreb, Rooseveltov trg 6, 10000, Zagreb, Croatia 2CEMAS – Centre for Estuarine and Coastal Marine Science, Department of Environmental Sciences, Informatics and Statistics, Universita Ca’ Foscari Venezia, Campo della Celestia, Castello 2737/b, 30122, Venice, Italy

Received 12 June 2015; revised 25 August 2015; accepted for publication 25 August 2015

The aim of this study was to describe the vocal repertoire of the Ponto-Caspian goby Neogobius fluviatilis and to compare the acoustic properties of this species with those of other soniferous Mediterranean gobies belonging to the Gobius lineage. Vocalizations and associated behaviours were recorded under controlled aquarium conditions in female and male N. fluviatilis. Sound emission was elicited by means of ‘intruder tests’, using an individual of the same or opposite sex as an intruder, and recording sounds using a hydrophone placed 20 cm from the shelter used as a nest for the resident fish. Five acoustic properties, including spectral and temporal properties, were measured from 13 individuals. The vocal repertoire of the species consisted of sequences of short vocalizations during both agonistic and reproductive intraspecific interactions. The wave form of each sound resolved in a pure sine wave composed of rapidly repeated pulses. Sounds lasted about 200 ms, showing an average fundamental frequency of about 80 Hz. Sound properties did not differ between reproductive and the aggressive contexts, and the general structure of sounds was highly stereotyped. The individual means of three acoustic independent traits characterizing the sounds of seven species of the Gobius lineage, including N. fluviatilis, were then entered in a discriminant function analysis to assess how well species could be differentiated on the basis of acoustics, and their degree of affinities. The results suggested that the pulse repetition rate of the sounds, i.e. the relative tonal/pulsatile nature of the sounds, was the most important property in differentiating the species, and that this trait may contain a high level of phylogenetic signal, as the species producing tonal sounds clustered together, in line with the results of recent molecular phylogenic studies. The results were discussed in light of the geological and phylogeographical events believed to have driven the diversification of European gobies. © 2015 The Linnean Society of London, Biological Journal of the Linnean Society, 2016, 117, 564–573.

ADDITIONAL KEYWORDS: acoustic properties – gobies – Messinian salinity crisis.

INTRODUCTION 2013). Among the vertebrates, the largest diversity of sound-generating mechanisms for acoustic commu- Acoustic signals convey crucial information on spe- nication has evolved in fishes (Myrberg & Lugli, cies, sex and individual identity, individual motiva- 2006; Fine & Parmentier, 2015). In comparison with tion and quality (Bradbury & Vehrencamp, 1998). tetrapods, fish have relatively simple central and The degree of similarity among acoustic signals in peripheral vocal mechanisms and thus typically lack groups of closely related species could be related to the ability to emit complex frequency-modulated calls phylogenetic relationships, as shown for anurans, (Rice & Bass, 2009). Sound production in a commu- insects, birds and mammals (Robillard et al., 2006; nicative context has been documented in over 800 Tavares et al., 2006; Cap et al., 2008; Gingras et al., fish species representing 109 families (Kasumyan, 2008), although phylogenetic reconstruction based on *Corresponding author. E-mail: [email protected] acoustics has seldom been attempted, and the state

564 © 2015 The Linnean Society of London, Biological Journal of the Linnean Society, 2016, 117, 564–573 SOUND PRODUCTION IN THE GOBIUS LINEAGE 565 of knowledge is at an early stage as compared with the exception of Odontobutis obscura, a basal gobioid amphibians or other animal groups. Nevertheless, as whose vocal repertoire was documented by Take- for anurans, the acoustic signals of teleosts could be mura (1984), most of the documented soniferous spe- useful for phylogenetic reconstruction, due to the cies are concentrated in the derived family Gobiidae. instinctual and stereotyped nature of their Within this family, sound production is mostly docu- vocalizations. Attempts have been made to study the mented in gobies occurring across the Mediter- species-specificity of acoustic signals and phylogenic ranean–Atlantic and Ponto-Caspian areas. These relationships and similarities in sound structure in species can be subdivided into two main lineages three teleost families in particular – batracoidids, according to the most recent molecular phylogenies pomacentrids and gobids (Malavasi, Collatuzzo & (Table 1, Thacker & Roje, 2011; Agorreta et al., Torricelli, 2008; Parmentier et al., 2009; Rice & Bass, 2013). The acoustic structure showed great variabil- 2009). In some cases, constraints related to body size ity, from pure tonal to pulsatile and complex sounds, seem to mask the species-specificity of the sounds, within the Gobius lineage, while only pulsatile and sounds do not succeed in discriminating species, sounds were reported for the Pomatoschistus lineage as was shown for clownfishes (Colleye & Parmentier, (Ladich & Kratochvil, 1989; Rollo et al., 2007; Mala- 2012). vasi et al., 2008; Sebastianutto et al., 2008; Polgar Sound production has been documented in at least et al., 2011; Amorim & Neves, 2007; Parmentier 21 gobioid species belonging to ten different genera et al., 2013; Table 1). A comparative analysis of the (Bass & McKibben, 2003; Polgar et al., 2011). Within sound structure within Mediterranean–Atlantic this large and diverse family, pulsatile and tonal gobies showed a clear distinction between these two sounds are emitted by the male as a part of the lineages, suggesting a good degree of congruence breeding and aggressive behavioural repertoire (Tor- between acoustic affinities and phylogenetic relation- ricelli, Lugli & Pavan, 1990; Lugli et al., 1997; Lugli ships (Malavasi et al., 2008). Given that the phy- & Torricelli, 1999; Malavasi et al., 2003; Myrberg & logeny of this group of species was probably driven Lugli, 2006; Amorim & Neves, 2007). Gobies are dis- by complex geological events related to separation of tributed worldwide, in marine, estuarine and fresh- the Tethys and Paratethys, the Messinian salinity water habitats. Acoustic communication has been crisis, and the subsequent re-flooding of the Atlantic widely investigated, especially in Mediterranean gob- Sea (Penzo et al., 1998; Huyse, Van Houdt & ies, such as in the genera Gobius, Padogobius, Zos- Volckaert, 2004; Miller, 2004; Malavasi et al., 2012; terisessor, Pomatoschistus and Knipowitschia (Bass Vanhove et al., 2012), the clarification of the phyloge- & McKibben, 2003; Myrberg & Lugli, 2006). Accord- netic relationships within these two lineages is of ing to existing molecular phylogenies, gobioid fishes particular phylogeographical interest. Nevertheless, (Gobiiformes, Gobioidei) can be subdivided into due to the systematic complexity of this group and several lineages (Agorreta et al., 2013), with some the high number of species, a great deal of informa- basal groups well resolved, such as Odontobutidae, tion is still required to obtain a reliable, complete Butidae and Eleotridae (Thacker, 2003, 2009; picture. Furthermore, the sound production mecha- Neilson & Stepien, 2009; Agorreta et al., 2013). With nism remains to be elucidated, at least with some

Table 1. Systematic position, geographical range and sound structure of the European soniferous gobies (Gobioidei, Gobiidae), according to the current literature (see text for references)

Ponto-Caspian Sub-families Lineages Atlantic–Mediterranean species species Sound structure

Gobiine-like Gobius Gobius niger, Proterorhinus Tonal, pulsatile and G. paganellus, marmoratus, complex (sounds composed G. cobitis, Neogobius of both pulsatile and tonal G. cruentatus, Zosterisessor melanostomus segments) ophiocephalus, Padogobius bonelli, P. nigricans Gobionelline-like Pomatoschistus Pomatoschistus marmoratus, Pulsatile P. pictus, P. minutus, P. microps, P. canestrinii, Knipowitschia panizzae K. punctatissima

© 2015 The Linnean Society of London, Biological Journal of the Linnean Society, 2016, 117, 564–573 566 S. HORVATIC ET AL. degree of detail, despite recent insight provided by channel system Kupa-Kupcina in Croatia (GPS coor- Parmentier et al. (2013), who suggested a cranio-pec- dinates; x, 45:31:53.7; y, 15:47:18.5; z, 93.2 m), where toral mechanism in Gobius paganellus. this species has the status of an invasive species. Recent molecular phylogenies reveal that Ponto- Fish were collected using electrofishing (electric unit, Caspian gobies form part of the Gobius lineage power: 7.5 kW) and injury and fatalities were care- (Agorreta et al., 2013), together with other Mediter- fully avoided. After capture, the fish were trans- ranean species, mostly belonging to the genus Go- ported in aerated containers from Croatia to the bius. With the exception of few sound recordings laboratory of the Ca Foscari University (Venice, reported for Proterorhinus marmoratus and Neogob- Italy) where they were housed in suitable aquaria. ius melanostomus, the vocal repertoire of these spe- Upon their arrival at the laboratory, fish were exam- cies remains to be described. Rollo et al. (2007) ined for sex on the basis of urogenital papilla (Miller, recorded a single sound emitted by Neogobius 1984). Larger fish used in the experiments were melanostomus in the field and then, using that single placed in smaller glass-tanks (capacity 120 L), each sound as a stimulus, recorded several other sounds individually, while the smaller individuals were in the laboratory. The aim of this study was to placed in larger glass-tanks (capacity 300 L). Shel- describe the vocal repertoire of Neogobius fluviatilis ters made of tiles were placed in each tank to serve (Pallas, 1814) and to compare the acoustic properties as a nest. Tanks were sound-proofed from surround- of this species with those of the other soniferous spe- ing noise using foam rubber shims as a base. Each cies belonging to the ‘Gobius lineage’ (sensu Agorreta tank was provided with a filtration system, proper et al., 2013), using the data of the present paper and substrate of a 5- to 10-cm-thick layer of sand on the those provided by Malavasi et al. (2008) for Mediter- bottom, and aerators to maintain oxygen levels. ranean species. The final objective was to explore the Water temperature (range 19–21 °C) and salinity degree to which the affinities in acoustic signals (0.05 PSU, obtained from tap water) were main- could be related to phylogenetic relationships within tained at levels within the average values found in this group of closely related species. the natural environment at the same time of the year. Fishes were left to acclimate in the tanks for 1 week and were fed chironomid larvae and mytili ad libitum. METHODS STUDY SPECIES Neogobius fluviatilis (Pallas, 1814), the monkey goby, SOUND COLLECTION belongs to a group of transverse gobiids that includes Experiments were set up to elicit sounds in male– the euryhaline Mediterranean Zosterisessor, exclu- male or female–female aggressive interactions (ag- sively freshwater Mediterranean Padogobius, and gressive behaviour) and male–female (courtship or four Ponto-Caspian groups, namely Mesogobius, spawning behaviour) reproductive interactions. Neogobius, Proterorhinus and Chasar, and the tad- According to the protocol used by Malavasi et al. pole gobies. The Sarmatic taxa (Neogobius, Mesogob- (2008), isolated resident fish (either males or ius, Proterorhinus and the tadpole gobies) together females) were placed in experimental tanks (120 L) with the Mediterranean Padogobius are synapomor- and left to establish territorial behaviour for about phic with the euryhaline Mediterranean Zosterises- 5–8 days before the beginning of recordings. Each sor ophiocephalus in reduction or loss of free pectoral isolated fish was provided with a tile shelter rays (Miller, 2004). The monkey goby is a euryhaline (10 9 20 cm). Before recordings, an intruder (either species, found on sandy bottoms, and distributed in male or female) was placed in a metal cage in front rivers and estuaries in the Ponto-Caspian basin. of the shelter entrance in order to elicit behavioural Spawning commences at the end of April, when tem- and acoustic responses. As regards reproductive peratures rise to about 13 °C, and reaches a peak in interactions, only male fish were used as the resident May, at temperatures of 18–19 °C. The majority of fish, assuming that only males are vocal during spawning is completed in June, with only limited courtship, according to the literature data on sonifer- occurrence in July (Miller, 2004). ous gobies. The lights and all pumping devices were switched off 10–15 min before the beginning of the experiments to minimize external noise. After the FISH COLLECTION AND HOUSING experiments, water temperature was measured with The experiment was conducted from the end of a digital thermometer, and the body size of each March until the beginning of June 2014, during the tested individual was measured using callipers. reproductive season of the studied species. Fish were Sounds were recorded with a custom-made hydro- caught during March and April 2014 in the artificial phone (Gulton Industries; sensibility: À164 dB, re

1 µPa, frequency response Æ 3 dB from 30 Hz to between-individual (CVb) coefficients of variation. 2 kHz) positioned about 20 cm from the entrance of The CVb/CVw ratio was used as a measure of the the nest, according to Malavasi et al. (2008). The relative variability among individuals. To test for hydrophone was connected to a conditioning ampli- the differences in the acoustic properties of males fier (B&K 2626) and sounds for analysis were between reproductive and aggressive contexts, and recorded using a portable digital audio recorder (Tas- between males and females within the aggressive cam Linear PCM Recorder; wav 16/44.1 k mono), contexts, a t-test was performed on log-transformed and stored on the recorder memory card, and were data. Individual means of acoustic properties were then imported to a personal PC. Kraun headphones related to size and temperature using the Pearson were connected to the digital audio recorder for bet- correlation to check for the dependency of sound ter sound quality. Overall, sounds were recorded structure on these two physical characteristics. To from 13 individuals (eight males, five females). Of assess the affinities of Neogobius sounds with the the eight males, four were tested with a male intru- other soniferous species of the ‘Gobius lineage’ der (aggressive interactions) and four with a mature (sensu Agorreta et al., 2013), recordings were female intruder (reproductive interactions). Each obtained from Malavasi et al. (2008) for six species recording lasted 30 min from the emission of the first belonging to the genera Gobius, Padogobius and Zos- sound. terisessor, and then analysed. To explore the affinities in terms of sound structure within this group of species, a multivariate statistical approach was used. SIGNAL PROCESSING AND SOUND ANALYSIS First, the log-transformed individual means of the All sounds were analysed in real time using Avisoft- five acoustic properties (DUR, NP, PRR, PF, FM) SASLab Pro Software. Analog signals were digital- recorded for the seven species investigated (with one ized at 1500 Hz sampling and were band-pass species, Padogobius bonelli, being split in the pul- filtered (band: 50–500 Hz) to eliminate acoustic com- satile and tonal components of the complex sounds, ponents different from those of fishes and which according to Malavasi et al., 2008) were tested for might disturb or distort the waveform of any fish correlation using the Pearson correlation. Then, a sound. The temporal and spectral structures of discriminant function analysis (DFA) was performed sounds were investigated using the time signal and on the log-transformed individual means of those power spectrum features provided by the software. A acoustic properties that proved to be independent on spectrogram of each sound was obtained by setting the basis of the correlations (DUR, PRR, PF). The the spectral parameters to achieve the best represen- aim was to assess how well individuals can be classi- tation of signals in relation to their acoustic struc- fied into the correct species, which acoustic proper- ture (window type: hamming; fast Fourier transform: ties differentiate the species, and their relative 256; frame: 100; band-width: 8 Hz; resolution: 5 Hz; distance and affinities based on acoustics. Significant overlap: 93.75%).The following acoustic properties differences among the species were examined by were measured following Malavasi et al. (2008): (1) quantifying Mahalanobis distances between the duration (DUR; total length of the call, measured in group centroids. milliseconds); (2) number of pulses (NP); (3) pulse repetition rate (PRR; obtained by dividing NP by DUR and expressed in Hz); (4) peak frequency (PF; RESULTS obtained from the power spectrum function); and (5) frequency modulation (FM; calculated as the differ- SOUND PRODUCTION IN NEOGOBIUS FLUVIATILIS AND ence between final PRR and initial PRR and INTRASPECIFIC VARIATION expressed in hertz). Neogobius fluviatilis emits sequences of short vocalizations during both agonistic and reproductive intraspecific interactions. The wave form of each STATISTICAL ANALYSIS sound resolves in a pure sine wave composed of To assess the level of intraspecific variation in sound rapidly repeated pulses, which represents a periodi- production, coefficients of variation [CV = (SD/ cally repeated longitudinal sound wave (Fig. 1). The X) 9 1000] were calculated for each property, on power spectrum revealed that most of the sound untransformed data, at each source of variation energy was concentrated within the narrow band of (within-individual and between-individual variation). 60–100 Hz (Fig. 1), with an average peak frequency Acoustic properties measured for each sound, indi- of about 80 Hz (Table 2). These properties revealed vidual means of acoustic properties and grand the tonal nature of the sound that can be assessed means calculated on the 13 individuals recorded on the basis of spectrogram, power spectrum were used to calculate within-individual (CVw) and and wave form of each sound (Fig. 1). Each sound

Figure 1. Structural details of two consecutive, representative sounds of Neogobius ﬂuviatilis: below, spectrogram (right) and power spectrum (left); above, the wave form of both sounds.

Table 2. Individual means (Æ SD) of body size, water quency slightly less than 100 Hz (Fig. 1). The single temperature of recordings, and the five measured acoustic sound is part of the sequence in which the intra- properties in each sex and behavioural context sound period (period between the completion of one sound and the onset of the next) lasted from 1.3 to Behavioural context Male (N = 8) Female (N = 5) 1.6 s (Fig. 1). Sounds lasted from 127 to 226 ms, with an average of 169.3 ms. The first two-thirds of Body size (mm) the sound duration was marked by a moderate Aggressive 132.7 Æ 8.3 116.2 Æ 11.8 increase in amplitude, which then rapidly declined Reproductive 128.5 Æ 11.2 – (Fig. 1). The average value of FM ranged from 2 to Temperature (°C) 13 Hz, indicating a slightly upward modulation. The Aggressive 20.2 Æ 0.4 19.8 Æ 0.4 number of pulses for each sound varied between Reproductive 20 Æ 0 – DUR (ms) eight and 16, with an average of 12.3 (Table 2). The Aggressive 179 Æ 29.5 163.62 Æ 20.5 pulse repetition rate ranged from 69 to 81 Hz, with Reproductive 161 Æ 38.5 – an average of 73.6 Hz (Table 2), overlapping with NP the mean peak frequency of 77 Hz, and constituting Aggressive 12.7 Æ 1.3 12.2 Æ 1.3 the fundamental frequency of the sound, again Reproductive 12 Æ 2.9 – revealing the harmonic nature of these vocalizations. PRR (Hz) The general structure of vocalizations was highly Aggressive 72.6 Æ 8.8 74.4 Æ 2.4 invariant both with respect to the behavioural con- Reproductive 74.8 Æ 1.0 – text (reproductive vs. aggressive context) and PF (Hz) between the sexes, and a single type of sound was Aggressive 77.2 Æ 6.6 78.4 Æ 2.4 recorded during all experimental trials, indicating Reproductive 78.8 Æ 3.6 – the stereotyped nature of the sound production. Of FM (Hz) the five acoustic properties of tonal segments, DUR, Aggressive 9.4 Æ 2.1 11.8 Æ 3.5 NP and FM (within reproductive interactions) and Reproductive 8.3 Æ 5.6 PRR (within aggressive interactions) had a CVb/CVw ratio > 1 (Table 3), suggesting a higher level of inter- DUR, duration; NP, number of pulses; PRR, pulse repeti- than intra-individual variability. This confirms the tion rate; PF, peak frequency (fundamental frequency); stereotyped nature of vocalizations. FM, frequency modulation. The fish produced sounds while swimming or while resting on the substrate. Sound emission of the is resolved in a main single harmonic and in a sounds was preceded by a clearly visible upward sinusoidal wave form, with most of the energy of the thrusting of the head, during which a dorsolateral sound concentrated at a single fundamental fre- movement of the opercula was performed. Breeding

Table 3. Intra-individual and inter-individual CV and Table 4. Pearson correlation coefficients of the relation- their ratio for the five acoustic properties measured, in ships between the individual means of the five acoustic each behavioural context properties of the goby species belonging to the Gobius lineage (N = 52) CVw CVb CVb/CVw Acoustic NP PRR PF FM property Aggr. Repr. Aggr. Repr. Aggr. Repr. DUR 0.54 * À0.15 0.12 0.22 DUR (ms) 0.20 0.15 0.16 0.23 0.81 1.51 NP 0.74 * 0.04 0.33 * NP 0.21 0.17 0.11 0.24 0.54 1.41 PRR À0.04 0.22 PRR (Hz) 0.06 0.06 0.08 0.01 1.25 0.21 PF 0.29 * PF (Hz) 0.09 0.47 0.06 0.34 0.63 0.71 FM 0.91 0.23 0.37 0.87 0.40 3.69 DUR, duration; NP, number of pulses; PRR, pulse repetition rate; PF, peak frequency (fundamental frequency); DUR, duration; NP, number of pulses; PRR, pulse repeti- FM, frequency modulation. tion rate; PF, peak frequency (fundamental frequency); *P < 0.05. FM, frequency modulation (absolute numbers, as in this case the value could be negative). axis, characterized by pulsatile sounds. The first discriminant function, as revealed by the standard males significantly changed their coloration, becoming structure coefficients of DFA (Table 5), differentiated gradually darker, and eventually black. In contrast species with a high PRR, and low DUR and PF (right to breeding males, females retained their semi-trans- side of the axis, tonal nature of the sounds, Fig. 2), parent, very pale coloration. Although individuals from those with low PRR and high DUR and PF (left were not free to interact with one another due to the side of the diagram, pulsatile sounds, Fig. 2). The cage, the change in coloration allowed for discrimina- second discriminant function accounted for 30.1% of tion between aggressive and courtship interactions. the variation, and further supported the separation An intense change of males towards darker col- of the tonal vs. pulsatile components, with DUR, oration was observed only during male–female inter- PRR and PF giving a negative contribution on this actions, making this a reliable indicator of courtship axis, and again with the major contribution given by motivation. PRR, followed by PF and DUR (Table 5). Neogobius There were no statistically significant differences fluviatilis clustered into the top right quadrant of in any of the behavioural properties, either between the chart; i.e. in comparison with the remaining six behavioural contexts or between sexes (t-test, species, this species is characterized by a high PRR p > 0.05). There were no statistically significant cor- and low PF. relations between any of the acoustic properties with temperature or with body size (Pearson correlation, N = 13, P > 0.05). DISCUSSION The results of the present study showed that intra- ACOUSTIC AFFINITIES WITHIN THE GOBIUS LINEAGE male and intra-female aggression and male courtship Results of the correlations showed that NP was sig- are associated with sound production in the Ponto- nificantly related to PRR and DUR, and PF and FM Caspian goby species Neogobius fluviatilis. Sound were also in turn significantly associated. Therefore, production consisted of short (about 200 ms), low-fre- only PRR, DUR and PF were used in the DFA quency (about 70–80 Hz) vocalizations, with a pure (Table 4). Using species as the grouping variable in tonal structure and highly stereotyped nature. the DFA, individuals were well classified into the Sounds were invariant, in their structure and mean correct species, with an average correct species clas- properties, in respect to both sex and behavioural sification rate of 88.46%. The first discriminant func- context. Temperature did not affect any acoustic tion accounted for 65.4% of the variation and property, probably due to the restricted range of distinguished three groups of species (Fig. 2): (1) Go- water temperature values recorded during the trials bius paganellus, the tonal component of Padogobius (19–20 °C). bonelli, Neogobius fluviatilis and Padogobius nigri- Sound production has been also documented in cans, on the positive side of the axis; (2) Gobius cobi- Neogobius melanostomus, with some calls recorded tis, the pulsatile component of Padogobius bonelli from individuals of the invading populations in the and Gobius niger in the central zone of the diagram,; United States (Rollo et al., 2007). These authors des- and (3) Zosterisessor ophiocephalus to the left of the ignated these calls as ‘pulse series’, although the

that species, both males and females produced tonal G. cobitis G. niger sounds during aggression, while males also emitted G.paganellus the same type of sounds during courtship. These N. fluviatilis affinities are expected, considering that both Neogob- P. bonelli P. bonelli ius and Proterorhinus share the Ponto-Caspian geo- P. nigricans graphical distribution area, and belong to the same Z. ophiocephalus clade (Simonovic, 1999; Medvedev et al., 2013). More surprising is the overlap in the structure and the acoustic properties of N. fluviatilis sounds with those documented for the Arno goby Padogobius nigricans (Lugli et al., 1995; Malavasi et al., 2008), which is an endemism of the Tuscan–Latium freshwater fish fauna of central Italy (Miller, 2004). The frequency, duration, general structure and stereotype of sound production and associated behaviours are strongly similar (Lugli et al., 1995), as the wave form and the power spectrum mostly overlapped, despite slight dif- Figure 2. Bivariate comparison between discriminant ferences in the mean values of the acoustic proper- functions DF1 and DF2 where individuals are delineated ties (present paper; Lugli et al., 1995). This high by species (different symbols represent different species). The dashed line separates species producing sounds with degree of affinity is consistent with the morphologi- a high pulse repetition rate (tonal sounds, right side) cal affinities between the two species, noted by Miller from species producing sounds with a low pulse repetition (2004). According to Miller, the Ponto-Caspian group rate (pulsatile sounds, left side). is synapomorphic with the Italian freshwater Pado- gobius in possessing a higher number of vertebrae – Table 5. Results of the discriminant function analysis (29 35), but also having a reduced or atrophied with species as the grouping variable and the three inde- swimming bladder, and reduced scapula. The most pendent acoustic properties as dependent variables recent molecular phylogenies agree in clustering the Ponto-Caspian gobies with other Mediterranean Standardized Factor struc- gobies belonging to the genera Gobius, Padogobius, coefficients ture coefficients Zosterisessor and Zebrus (Thacker & Roje, 2011; Agorreta et al., 2013; Fig. 3). The results of the pre- Acoustic property DF1 DF2 DF1 DF2 sent study support these phylogenies on an acoustic basis, as Neogobius clustered with both the Padogob- À À À À DUR 0.39 0.62 0.11 0.15 ius species and Gobius paganellus, i.e. species pro- PRR 0.90 À0.55 À0.67 0.20 ducing vocalizations with a high pulse repetition PF À0.89 À0.77 À0.38 0.64 rate, that result in an acoustic tonal structure, and The results shown are the standardized canonical coeffi- the main separation among these species, primarily cients and the factor structure coefficients. Factor struc- the tonal component of Padogobius bonelli, was due ture coefficients are the bivariate correlation between a to differences in peak frequency. In contrast, the species/individual value for a dependent variable and the comparative analysis conducted here revealed that individual’s discriminant function score and are thus Zosterisessor ophiocephalus and Gobius niger formed important to characterize the contribution of each depen- a separate cluster on an acoustic basis, due to the dent variable. lower pulse repetition rate and the consequent pulsatile nature of their sounds. This acoustic affinity wave forms of these sounds were not expanded and parallels the molecular affinities revealed by recent we strongly suspect that these authors did not per- phylogenic studies, as these two species clustered ceive the tonal nature of these calls. If this is the together as a sister group (Huyse et al., 2004; case, the sounds of N. melanostomus would be simi- Thacker & Roje, 2011; Agorreta et al., 2013; Fig. 3). lar to the con-generic species recorded by the present Thus, our results suggest strongly that the pulse study. Regardless, the vocal repertoire of N. me- repetition rate, i.e. the degree to which pulses are lanostomus requires more detailed study, considering stacked together in producing a continuum from pul- the importance of this species as an alien invader. satile to tonal sounds, is the most important property Furthermore, sounds recorded from the Ponto-Cas- in species discrimination, and that this acoustic pian goby Proterorhinus marmoratus (Ladich & Kra- property could contain a certain degree of a phyloge- tochvil, 1989) appeared to have a very similar netic signal. A robust phylogenetic analysis based on structure to those described here for N. fluviatilis.In goby acoustics was beyond the scope of the present

Figure 3. Two clusters extracted and adapted from Thacker & Roje (2011) (A) and Huyse et al. (2004) (B) showing the relationships between certain soniferous gobies, and the wave forms of their typical sounds. analysis, due to the lack of intervening species in the (2013), conducted on Gobius paganellus, which available molecular phylogenies, and an exact paral- suggested that sound might be generated by the lelism between documented acoustic production and periodic contraction of the levator pectoralis muscle. resolved systematic position of the different species. This study, together with the observations reported Nevertheless, both the present study and the anal- by Lugli & Fine (2003), suggests that the swimblad- ysis by Malavasi et al. (2008), conducted at a higher der is not involved in sound emission. Note that both phylogenetic scale in comparing the Gobius and the N. fluviatilis and Padogobius nigricans, i.e. gobies Pomatoschistus lineages (sensu Agorreta et al., 2013) producing pure tonal sounds, lack a swimbladder. By on an acoustic basis, seem to indicate that affinities contrast, many other goby species have retained the in sound production, specifically in the pulse repeti- swimbladder, which poses an interesting question tion rate, correlate with phylogenetic relationships. about the evolutionary pattern of swimbladder loss Alternative hypotheses are related to convergent evo- or retention in this group. In light of these considera- lution or rapid evolution due to sexual selection, as tions, a comparative analysis of the sound emission observed for vocalizing groups of tetrapods, espe- mechanism in gobies is urgently needed to depict the cially anurans (Robillard et al., 2006; Tavares et al., phylogeny of acoustic communication in this fish 2006; Cap et al., 2008; Gingras et al., 2013). To dis- group. Assuming that the acoustic affinities relate to criminate the contribution of phylogenetic signal phylogeny in the Gobius lineage, certain preliminary from that of evolutionary convergence, analyses hypotheses on the natural history of this lineage can relating the molecular affinities and acoustic affini- be presented. ties in the same group of species are needed, as well The acoustic clustering of N. fluviatilis with other as a full clarification of the emission mechanisms in species producing tonal sounds, i.e. the Padogobius the different species. The only experimental study on complex and Gobius paganellus, in turn clustered a sound emission mechanism available for a goby together according to the molecular phylogenies pro- species is that recently provided by Parmentier et al. vided by Penzo et al. (1998) and Huyse et al. (2004),

© 2015 The Linnean Society of London, Biological Journal of the Linnean Society, 2016, 117, 564–573 572 S. HORVATIC ET AL. suggest a common ancestor for this group of species. Bass AH, McKibben JR. 2003. Neural mechanisms and According to Penzo et al. (1998), a possible hypothe- behaviors for acoustic communication in teleost fish. Pro- sis is a Ponto-Caspian ancestor that migrated gress in Neurobiology 69: 1–26. through the Pannonian channel to the Mediter- Bradbury JW, Vehrencamp SL. 1998. Principles of animal ranean basin in the early/middle Miocene (24– communication. Sunderland, MA: Sinauer Associates. 15 Mya). According to this scenario, during the Mes- Cap H, Deleporte P, Joachim J, Reby D. 2008. Male – sinian salinity crisis (5.5 Mya), when virtually the vocal behavior and phylogeny in deer. Cladistics 24: 917 entire Mediterranean Sea was desiccated, several 931. hyper- and hypo-saline lakes appeared (Huyse et al., Colleye O, Parmentier E. 2012. Overview of the diversity of sounds produced by the clownfishes (Pomacentridae): 2004). Those hypo-saline lakes became refuges for importance of acoustic signals in their peculiar way of life. the euryhaline species within them. Those species PLoS ONE 7: 1–11. had to adapt to the new environmental conditions, Fine M, Parmentier E. 2015. Mechanisms of fish sound i.e. a freshwater lifestyle. Such isolation probably production. In: Ladich F, ed. Sound communication in gave rise to the freshwater endemism existing today fishes. Berlin: Springer, 77–126. in the Mediterranean (Miller, 1990). With the open- Gingras B, Mohandesan E, Boko D, Fitch WT. 2013. ing of the Strait of Gibraltar (5.33 Mya), and subse- Phylogenetic signal in the acoustic parameters of the adver- quent re-flooding of the Mediterranean basin, the tisement calls of four clades of anurans. BMC Evolutionary Mediterranean again became a marine habitat. Due Biology 13: 134. to its distribution, it is possible that this common Huyse T, Van Houdt J, Volckaert FAM. 2004. Paleocli- ancestor had to adapt then to the new environment, matic history and vicariant speciation in the ‘sand goby’ i.e. a marine lifestyle. Adaptation to the new and group (Gobiidae, Teleostei). Molecular Phylogenetics and free ecological niches probably led to radiation Evolution 32: 324–336. resulting in the present-day goby fauna. A similar Kasumyan AO. 2008. Sounds and sound production in evolutionary Messinian and post-Messinian scenario fishes. Journal of Ichthyology 48: 981–1030. with an ancestral freshwater life style and a derived Ladich F, Kratochvil H. 1989. Sound production in the marine life style was proposed by Malavasi et al. Marmoreal goby Proterorhinus marmoratus (Pallas) (Gobi- (2012) for a group of Mediterranean sand gobies on idae: Teleostei). Zoologische Jahrbucher Abteilung fuer All- the basis of behavioural phylogeny. Although prelim- gemeine Zoologie und Physiologie der Tiere 93: 501–504. inary and, to certain degree still speculative, these Lugli M, Fine ML. 2003. Acoustic communication in two scenarios indicate that acoustic properties should not freshwater gobies: ambient noise and short-range propaga- be neglected when reconstructing phylogenetic tion in shallow streams. Journal of Acoustic Society of America 114: 512–521. pathways. Lugli M, Torricelli P. 1999. Prespawning sound production in Mediterranean sand-gobies. Journal of Fish Biology 54: 691–694. ACKNOWLEDGEMENTS Lugli M, Pavan G, Torricelli P, Bobbio L. 1995. Spawn- ing vocalizations in male freshwater gobiids (Pisces: Gobi- We are grateful to Dr Marco Lugli and Irene Bendaz- idae). Environmental Biology of Fishes 43: 219–231. zoli for their kind and useful assistance in sound Lugli M, Torricelli P, Pavan G, Mainardi D. 1997. Sound recording and analyses, and Linda Zanella for proof- production during courtship and spawning among freshwa- reading the manuscript. We are grateful to Professor ter gobiids (Pisces: Gobiidae). Marine and Freshwater Beha- Eric Parmentier and two anonymous referees for viour and Physiology 29: 109–126. their comments that contributed to improve the over- Malavasi S, Torricelli P, Lugli M, Pravoni F, Mainardi all quality of the manuscript. D. 2003. Male courtship sounds in a teleost with alternative reproductive tactics, the grass goby, Zosterisessor ophiocephalus. Environmental Biology of Fishes 66: 231–236. Malavasi S, Collatuzzo S, Torricelli P. 2008. Interspecific REFERENCES variation of acoustic signals in Mediterranean gobies (Perci- Agorreta A, San Mauro D, Schliewen U, Van Tassell JL, formes, Gobiidae): comparative analysis and evolutionary Kovaci c M, Zardoya R, Ruber€ L. 2013. Molecular phylo- outlook. Biological Journal of the Linnean Society 93: 763– genetics of Gobioidei and phylogenetic placement of Euro- 778. pean gobies. Molecular Phylogenetics and Evolution 69: Malavasi S, Gkenas C, Leonardos I, Torricelli P, Da M. 619–633. 2012. The phylogeny of a reduced ‘sand goby’ group based Amorim MCP, Neves ASM. 2007. Acoustic signalling on behavioural and life history characters. Zoological Jour- during courtship in the painted goby, Pomatoschistus nal of the Linnean Society 165: 916–924. pictus. Journal of the Marine Biological Association of the Medvedev DA, Sorokin PA, Vasil’ev VP, Chernova NV, United Kingdom 87: 1017–1023. Vasil’eva ED. 2013. Reconstruction of phylogenetic

relations of Ponto-Caspian gobies (Gobiidae, Perciformes) of behavioral sequences by direct optimization. Cladistics based on mitochondrial genome variation and some prob- 22: 602–633. lems of their taxonomy. Journal of Ichthyology 53: 702–712. Rollo A, Andraso G, Janssen J, Higgs D. 2007. Attraction Miller PJ. 1984. The tokology of gobioid fishes. In: Potts and localization of round goby (Neogobius melanostomus)to GW, Wootton RJ, eds. Fish reproduction: strategies and tac- conspecific calls. Behaviour 144: 1–21. tics. London: Academic Press, 119–153. Sebastianutto L, Picciulin M, Costantini M, Rocca M, Miller PJ. 1990. The endurance of endemism – the Mediter- Ferrero E. 2008. Four types of sounds from one winner: ranean fresh-water gobies and their prospects for survival. vocalizations during territorial behaviour in the red- Journal of Fish Biology 37: 145–156. mouthed goby Gobius cruentatus (Pisces, Gobiidae). Acta Miller PJ. 2004. Gobiidae I. In: Miller PJ, ed. The freshwa- Ethologica 11: 115–121. ter fishes of Europe. Vol. 1. Wiesbaden: Aula-Verlag, 157– Simonovic PD. 1999. Phylogenetic relationships of Ponto- 398. Caspian gobies and their relationship to the Atlantic– Myrberg AA, Lugli M. 2006. Reproductive behavior and Mediterranean Gobiidae. Journal of Fish Biology 54: 533– acoustical interactions. In: Ladich F, Collin SP, Moller P, 555. Kaapor BG, eds. Communication in fishes, vol. 1. Enfield, Takemura A. 1984. Acoustical behavior of the freshwater NH: Science Publishers, 146–176. goby Odontobutis obscura. Bulletin of the Japanese Society Neilson ME, Stepien CA. 2009. Escape from the Ponto-Cas- for the Science of Fish 50: 561–564. pian: evolution and biogeography of an endemic goby spe- Tavares ES, Baker AT, Pereira SL, Miyaki CY. 2006. cies flock (Benthophilinae: Gobiidae: Teleostei). Molecular Phylogenetic relationships and historical biogeography of Phylogenetics and Evolution 52: 84–102. neotropical parrots (Psittaciformes: Psittacidae: Arini) Parmentier E, Lecchini D, Frederich B, Christophe B, inferred from mitochondrial and nuclear DNA sequences. Mann D. 2009. Sound production in four damselfish (Das- Systematic Biology 55: 454–470. cyllus) species: phyletic relationships? Biological Journal of Thacker CE. 2003. Molecular phylogeny of the gobioid fishes the Linnean Society 97: 928–940. (Teleostei: Perciformes: Gobioidei). Molecular Phylogenetics Parmentier E, Kever L, Boyle K, Corbisier Y, Sawelew and Evolution 26: 354–368. L, Malavasi S. 2013. Sound production mechanism in Go- Thacker CE. 2009. Phylogeny of Gobioidei and placement bius paganellus (Gobiidae). The Journal of Experimental within Acanthomorpha, with a new classification and inves- Biology 216: 3189–3199. tigation of diversification and character evolution. Copeia 1: Penzo E, Gandolfi G, Bargelloni L, Colombo L, Patar- 93–104. nello T. 1998. Messinian salinity crisis and the origin of Thacker CE, Roje DM. 2011. Phylogeny of Gobiidae and freshwater lifestyle in western Mediterranean Gobies. identification of gobiid lineages. Systematics and Biodiver- Molecular Biology and Evolution 15: 1472–1480. sity 9: 229–347. Polgar G, Malavasi S, Cipolato G, Georgalas V, Clack Torricelli P, Lugli M, Pavan G. 1990. Analysis of sound JA, Torricelli P. 2011. Acoustic communication at the produced by male Padogobius martensi (Pisces: Gobiidae) water’s edge: evolutionary insights from a mudskipper. and factors affecting their structural properties. Bioacous- PLoS ONE 6: e21434. tics 2: 261–275. Rice AN, Bass AH. 2009. Novel vocal repertoire and paired Vanhove MPM, Economou AN, Zogaris S, Larmuseau swimbladders of the three-spined toadfish, Batrachomoeus MHD, Giakoumi S, Kalogianni E, Volckaert FAM, trispinosus: insights into the diversity of the Batrachoidi- Huyse T. 2012. Phylogenetics and biogeography of the Bal- dae. Journal of Experimental Biology 212: 1377–1391. kan ‘sand gobies’ (Teleostei: Gobiidae): vulnerable species in Robillard T, Legendre F, Desutter-Grandcolas L, need of taxonomic revision. Biological Journal of the Lin- Grandcolas P. 2006. Phylogenetic analysis and alignment nean Society 105: 73–91.

Comparative analysis of sound production between the bighead goby Ponticola kessleri and the round goby Neogobius melanostomus: Implications for phylogeny and systematics

Sven Horvatić & Lea Bem & Stefano Malavasi & Zoran Marčić & Ivana Buj & Perica Mustafić & Marko Ćaleta & Davor Zanella

Received: 8 February 2018 /Accepted: 3 March 2019 /Published online: 22 March 2019 # Springer Nature B.V. 2019

Abstract Divergence in acoustic traits between closely Some acoustic features (sound rate and duration) proved related species can be explained by phylogenetic histo- to be stereotyped and could be considered a species- ry. In gobies, phylogenies reconstructed with acoustic specific trait that could potentially be utilized to discrim- signals primarily overlap with studies based on morpho- inate males or used by females for mate assessment. The logical or molecular data. Here, sound production of the tonal sounds appear to have a deeper origin within the two Ponto-Caspian gobies, Neogobius melanostomus Benthophilinae subfamily, as all acoustically investigat- and Ponticola kessleri, was recorded in controlled con- ed species to date have been able to produce this sound ditions and compared to determine the degree of inter- type. The recorded vocal repertoire represents a baseline specific acoustic variation across benthophilin gobies. for future comparative acoustic studies among the Both species produced tonal-like sounds characterized benthophilin gobies, aiming to gain additional informa- by unique temporal and spectral properties during ago- tion on the evolution of acoustic communication within nistic and reproductive intraspecific interactions, while Ponto-Caspian gobies and highlighting their importance the acoustic comparison revealed that the vocalizations in reconstructing phylogenetic relationships. of these two species differ in almost every acoustic property. N. melanostomus vocal structure was Keywords Ponto-Caspian gobies . Acoustics . Tonal . characterised by short (c. 100 ms), low-frequency (< Phylogeny 100 Hz) tonal sounds repeated at a relatively faster rate, while P.kessleri sounds appeared as a broadband, downward frequency modulated longer calls (c. 450 ms). Introduction

Determining species identity can be an important task S. Horvatić : L. Bem : Z. Marčić : I. Buj : P. Mustafić : D. Zanella (*) for related species in sympatry (Zeyl et al. 2016). Spe- Department of Zoology, Faculty of Science, University of Zagreb, cies identity may be encoded by acoustic signals Rooseveltov trg 6, 10000 Zagreb, Croatia (Myrberg et al. 1978; Spanier, 1979;Crawford1997; e-mail: [email protected] Lobel 1998) and several examples demonstrating diver- S. Malavasi gence in acoustic traits between closely related species Department Environmental Sciences, Informatics and Statistics, can be found in different vocal taxa (Robillard et al. Ca` Foscari University of Venice, Via Torino 155, 30172 Venezia 2006; Cap et al. 2008; Gingras et al. 2013; Hamao et al. Mestre, Italy 2015). In many different groups of vocal animals, phy- M. Ćaleta logenies reconstructed with acoustic signals are congru- Faculty of Teacher Education, University of Zagreb, Savska cesta ent with studies based on morphological and molecular 77, 10000 Zagreb, Croatia data (Cocroft and Ryan 1995; Laiolo and Rolando 2003; 728 Environ Biol Fish (2019) 102:727–739

Robillard and Desutter-Grandcolas 2004; Malavasi et al. multiple nuclear and mitochondrial genes to indicate 2008; Rice and Bass 2009;Horvatić et al. 2015). In that Ponto-Caspian gobies are deeply nested within the fishes, the auditory system plays a crucial role in their BGobius^ lineage, together with other Mediterranean life history, enabling them to interpret information from genera, supporting their close relationship. Therefore, the acoustic environment, mainly for predator and prey while the systematic position and phylogenetic relation- detection (Remage-Healey and Bass 2006), and for ships of Ponto-Caspian gobies await clearer resolution, inter- and intraspecific communication (Bradbury and the goal of this study was to investigate the phylogenetic Vehrencamp 1998; Malavasi et al. 2008; Colleye et al. relationships within Benthophilinae species based on 2011;Amorimetal.2013; Mélotte et al. 2016). Due to the structure of their acoustic signals. The main objec- the instinctual and stereotyped nature of their vocaliza- tive was to record the sound production of two Ponto- tions, teleost sounds are useful for phylogenetic recon- Caspian gobies, the bighead goby Ponticola kessleri structions (Malavasi et al. 2008; Amorim et al. 2011). (Günther 1861) and the round goby Neogobius Within gobies (Gobiiformes; Gobiidae), drumming, melanostomus (Pallas, 1814) under controlled aquarium stridulatory and tonal sounds are emitted by the male conditions, and to quantitatively describe the spectral as a part of the breeding and aggressive behavioural and temporal properties of their acoustic signals. Al- repertoire (Torricelli et al. 1990;Luglietal.1997; though round goby sounds have previously been record- Lugli and Torricelli 1999; Malavasi et al. 2003; ed (Rollo et al. 2007), no detailed vocal repertoire of this Myrberg and Lugli 2006; Amorim and Neves 2007). species was available in the literature. A second goal In recent decades, the vocal repertoire has been de- was to compare the acoustic signals of these two species scribed for at least 23 species of gobies, primarily be- and to determine the degree of interspecific variation longing to the Atlantic and Mediterranean genera across benthophilin gobies, based on acoustic data. The Gobius, Padogobis, Zosterisessor, Knipowitschia and vocal repertioire of these two species may represent a Pomatoschistus (Lugli et al. 1997; Lugli et al. 2003; first insight into future phylogenetic comparisons within Malavasi et al. 2008; Malavasi et al. 2009;Lugli2010; the subfamily, and could provide additional information Zeyl et al. 2016). Within the Ponto-Caspian lineage, on the evolution of acoustic communication within the however, few species have been acoustically tested Ponto-Caspian gobies. (Neogobius melanostomus, N. fluviatilis and Proterorhinus marmoratus) (Ladich and Kratochvil 1989; Rollo et al. 2007;Horvatić et al. 2015), and the Materials and methods vocal repertoires of most species remain unknown. In addition, the systematic position and phylogenetic rela- Fish collection tionships of Ponto-Caspian gobies have been disputed and show certain unresolved aspects. Berg (1949)sep- Fifteen specimens of Ponticola kessleri (140–180, LT, arated the relict Sarmatian gobies from the genus Gobius mm, seven males and eight females) and 15 specimens into a single genus Neogobius. Over the years, the genus of Neogobius melanostomus (120–170, LT, mm, nine Neogobius proved to be paraphyletic, and recently, males and six females) were collected from the Drava Neilson and Stepien (2009) grouped all Ponto-Caspian River near Osijek, (GPS coordinates: 45°33′45.73’ N, gobies into a monophyletic lineage within the subfamily 18°42′1.37″E) and from the Sava River downstream of Benthophilinae, further dividing them into three well- Zagreb (GPS coordinates: 45°44′19.28^N, 16°13′ supported tribes: Neogobiini (genus Neogobius), 48.40″E), respectively. Individuals were caught using Ponticolini (genera Babka, Mesogobius, Ponticola,and electrofishing (electric unit, power: 7.5 kW) in spring Proterorhinus), and Benthophilini (genera Benthophilus 2016 and 2017 from artificial riprap systems. and Caspisoma). Furthermore, Thacker and Roje (2011) examined the molecular interrelationships between Fish housing and experimental procedure many gobiid lineages and recognized all the proposed tribes of benthophilin species from Neilson and Stepien Upon arriving to laboratory, fish were sexed on the basis (2009), though with a slightly different position of the of urogenital papilla (Miller 1984) and housed in suitable tribe Neogobiini, which they placed as a sister group to aquaria, either in experimental tanks or in the community the Ponticolini tribe. Later, Agorreta et al. (2013)used tank. All tank temperatures ranged from 16 to 22 °C Environ Biol Fish (2019) 102:727–739 729 throughout the observation period. Each tank was sensitivity: −165 re 1 VμPa−1, frequency response from equipped with a filtration system, 5–10 cm thick sand 2 Hz to 30 kHz) positioned about 10 cm from the layer and aerators to maintain proper oxygen levels. entrance of the nest to avoid signal attenuation and alter- Artificial shelters made of tiles (10 × 20 cm) were placed ation due to tank reflections and reverberation, according in each tank to serve as a nest. Experimental tanks (90 L) to Akamatsu et al. (2002). Acoustic signals for analysis were sound-proofed from surrounding noise using a thick were recorded using a portable digital audio recorder layer of acoustically isolating material (foam rubber (ZOOM H4n; wav 16/44.1 k mono), stored on the re- shims). The experiments were carried out during the corder memory card and monitored on headphones. An- reproductive season of both investigated species, from alog signals were digitalized at 4000 Hz (16 bit) and March until August 2016 and 2017, and outside the analysed using Avisoft-SASLab Pro 5.2 Software (512- reproductive season following the protocol proposed by point fast fourrier transform FFT, Hamming window). Malavasi et al. (2008). The BIntruder test^ was set up to All sounds were band-pass filtered (band: 50–500 Hz) to elicit sounds in male–male, female–female or female– eliminate background noise and + 10 dB amplification male aggressive interactions (aggressive behaviour) and was applied for better auditory and visual inspection of male–female (courtship or spawning behaviour) repro- the audio tracks. Only sounds with a higher signal to ductive interactions. For the experiment, the tested resi- noise ratio were analysed. Temporal features were mea- dent fish (either male or female) was first measured (LT) sured from the oscillogram while frequency variables with a ruler and then placed indivdually into experimen- were obtained from the power spectra (filter bandwidth tal tanks (90 L) and left to establish territorial behaviour 300 Hz, 256 points FFT, frame size 100%, time overlap for about three days before the beginning of recordings. 96.87% and a Hamming window). The following acous- The remaining individuals from both species were sepa- tic properties were measured: (1) sound rate (SR; sounds rated by sex, and placed in four larger community glass- min−1); (2) duration (DUR; milliseconds); (3) number of tanks (capacity 300 L) to be used as the intruder during pulses (NP); (4) pulse repetition rate (PRR; obtained by the experiments. Lights and all pumping devices were dividing NP by DUR and expressed in Hz); (5) peak switched off 10–15 min before the beginning of experi- frequency (PF; Hz); (6) frequency modulation (FM; after ments to minimize external noise. Prior to the start of the sound has been divided on three sections, frequency recording, an intruder (either of the same or opposite sex, modulation was calculated as the difference between final depending on the interaction being tested) was placed in PRR and initial PRR and expressed in Hz). a metal cage and lowered into the tank and positioned in front of the shelter entrance in order to elicit behavioural Data analysis and acoustic responses. Each audio recording lasted 30 min from the emission of the first sound and during We calculated the mean and standard deviation (x ± SD) the experiments, the acoustic signals were noted. After for each of the six acoustic variables emitted by individ- the experiments, water temperature was measured with a ual fish of both species and tested them for inter- and digital thermometer and the intruder fish was placed back intraspecific differences. Sound features were examined in the community tank, and the process was repeated using from five to ten sounds with high sound quality/ with two other individuals during one recording session. noise ratio for each subject. Since the assumption of Once the resident had fought against all the other in- normality was not met for all acoustic parameters (Kol- truders, it was replaced in the experimental tank by mogorov-Smirnov test, P < 0.05), non-parametric another fish. Each individual was tested acoustically at Kruskal-Wallis ANOVA by Ranks followed by subse- least twice, though in inverted roles (resident vs intruder), quent post hoc multiple comparison test was used to waiting at least five days between interactions, to avoid investigate variation of acoustic properties between indi- the effects of the previous outcome (Torricelli et al. 1988; viduals of each species (intraspecific variation). Differ- Sebastianutto et al. 2008). ences in the mean acoustic properties of vocal subjects between different behaviour contexts (Breproductive^ vs. Acoustic methods Baggressive) and species (N. melanostomus v. P. kessleri) were tested using the Mann-Whitney U test on untrans- Sounds were recorded using a High Tech Inc. hydro- formed data. Since the reproductive and aggressive phone with built-in preamplifier (HTI 94 SSQ, sounds produced by N. melanostomus did not differ 730 Environ Biol Fish (2019) 102:727–739

(Mann-Whitney U test, P > 0.05), the mean values were (Table 2). Moreover, pulse repetition rate corresponded pooled together and used as a single set of acoustic to peak frequency and ranged from 55.9 to 124.8 Hz variables for further comparative analysis. In addition, (87.2 ± 5.9 Hz) (Table 2), suggesting that these sounds individual means of the six acoustic properties (SR, are purely tonal. Frequency modulation ranged from DUR, NP, PRR, PF, FM,) were tested against each other −49.1to56.3Hz(3.5±5.9Hz),indicatingslightupward for correlation using the Spearman rank correlations, and modulation of the sounds. Sound rate ranged from one to then subsequently related to size (LT) and temperature to 46 sounds per minute (10.1 ± 4.4), and it would speed up check for the dependency of sound structure on these two as the intruder approached the nest entrance (Table 2). physical characteristics. To examine individual acoustic The comparison between individual means of acoustic differences (stereotypy) among individuals of both spe- signals revealed that all sound features differed signifi- cies, the CVb/CVw ratio was examined. We calculated the cantly between N. melanostomus males (Kruskal-Wallis mean within-individual (CVw), and between-individual test: d.f. = 6 n = 88, all tests P < 0.05). From the six (CVb) coefficients of variation (CV=SD/x) × 1000; on acoustic features, DUR was highly and positively corre- untransformed data), utilizing overall means and SDs for lated with NP (rs = 0.8, P < 0.05) and negatively with each acoustic variable using the previously calculated PRR (rs = − 0.8, P < 0.05) and with PF (rs = −0.8, mean values for each subject. A ratio of >1.0 suggests P < 0.05). In addition, PF was positively correlated with that the acoustic parameter is more variable between PRR (rs =0.9,P < 0.05). There were no statistically sig- individuals relative to its variability within individuals, nificant correlations between acoustic properties and wa- and could be used as a cue for individual discrimination ter temperature or body size, due likely to the restricted and stereotypy (Christie et al. 2004). All statistical anal- range of values of these two physical parameters (Spear- yses were carried out with STATISTICA 13 (Statsoft man correlation, P > 0.05). Strong stereotypy was found Inc., Tulsa, OK, USA). Results were considered signifi- for SR, DUR, and NP, which were more variable among- cant at P <0.05. than within-individuals (CVb/CVw >1.0,Table3).

Ponticola kessleri Results All Ponticola kessleri males emitted sounds during the Neogobius melanostomus territory defence interactions, in an agonistic context, while no females produced any sounds during trials. All seven resident males of Neogobius melanostomus The same males were retested for sound production produced short vocalizations during both reproductive within a reproductive context, with a female intruder, and agonistic intraspecific interactions. No females pro- but none were vocal during courtship or pre-spawning duced sounds during the acoustic experiments, neither in trials. Acoustic signals were produced by both resident or the resident nor intruder role, while males in the intruder intruder fish and vocalizations were recorded from nine role also did not produce sounds. Three males were vocal males in total (Table 1). Sounds were produced soniferous during only one context (two reproductive, singly, or more commonly, in short sequences at a rela- one aggressive), while the remaining four individuals tively low rate, with the sound rate ranging from one to produced sounds during both reproductive and aggres- five (mean ± S.D.) (2.5 ± 0.7) (Table 2). Signals were sive interactions (Table 1). Each sound is part of a long relatively long in duration, which ranged from 189.1 to sequence and characterized by intra-sound amplitude 879.8 ms (457.9 ± 54.4 ms) and composed from 17 to 72 variation, which progressively increased throughout first pulses (44.9 ± 4.4), yielding a continuous tone-like wave- two-thirds of the sound duration, and then rapidly damp- form (Fig. 2,Table2). Energy ranged from 80 to 400 Hz ened (Fig. 1). Vocalizations were short and lasted from (Hz), exhibiting broadband nature of the sound with a 50.5 to 217.1 ms (mean ± S.D.) (114.8 ± 10.4 ms) com- strong first harmonic. Peak frequency was, in most cases, posed from five to 15 pulses (9.2 ± 0.4) (Table 2). Spec- the fundamental frequency, ranging from 81 to 209.9 Hz trographically, most of the sound energy appeared as a (104.9 ± 11.2 Hz) (Table 2), though in some cases, the single tone band within the range from 50 to 200 Hz spectrogram displayed two or three additional spectral (Hz), with a maximum amplitude represented at the peak harmonic components (< 400 Hz). Pulse repetition rate frequency, ranging from 52 to 128 Hz (86 ± 6.5 Hz) corresponded to peak frequency and ranged from 64.1 to Environ Biol Fish (2019) 102:727–739 731

Table 1 Total number of individuals acoustically tested and sounds analysed for each species with behavioural context (aggressive and reproductive) and experimental role (resident and intruder)

Male, TL (mm) Experimental role Sound production n. of sounds Temp. °C (resident/intruder) context recorded

Neogobius melanostomus #1, 161 Resident Reproductive 60 22.1 #2, 144 Resident Aggressive 6 22.0 Reproductive 13 22.2 #3, 125 Resident Aggressive 5 22.1 Reproductive 5 21.7 #4, 155 Resident Aggressive 27 21.8 Reproductive 14 21.2 #5, 145 Resident Aggressive 17 21.5 Reproductive 18 21.1 #6, 135 Resident Reproductive 43 19.1 #7, 154 Resident Aggressive 52 18.0 Ponticola kessleri #1, 146 Resident Aggressive 27 17.5 #2, 149 Resident Aggressive 26 19.3 #3, 147 Resident Aggressive 50 19.6 #4, 176 Intruder Aggressive 32 19.6 #5, 170 Resident Aggressive 34 20.9 Intruder Aggressive 13 #6, 171 Resident Aggressive 46 17.9 #7, 166 Intruder Aggressive 2 19.3 #8, 140 Intruder Aggressive 6 17.7 #9, 154 Resident Aggressive 12 22.3 Intruder Aggressive 27

Values of both water temperature (mean, °C) and emitter’s size (total length, mm) are reported. For each individual, from five to ten sounds, exhibiting good signal to noise (S/N) ratio, were analyzed and subsequently used in further inter- and intraspecific analysis. Individual #7 was subsequently excluded from inter- and intraspecific analysis due to small sample size, i.e. only two sounds recorded

128.8 Hz (99.9 ± 5.6 Hz), while the frequency modula- N. melanostomus vs. P. me ss le ri sounds tion indicated downward modulation of sounds and ranged from −39.2 to 38.3 Hz (−9.1 ± 6.8 Hz) The mean acoustic properties were compared between (Table 2). At the intraspecific level, all investigated sound the species, and SR, DUR, NP, PF and FM were found properties differed significantly across P. kessleri individ- to differ significantly between species (Mann-Whit- uals (Kruskal-Wallis test: d.f. = 7, n =75,all P < 0.05) ney U test, N1 =7,N2 =8, P < 0.05 all tests) while except for SR (Kruskal-Wallis test: d.f. = 7, n = 75, all PRR showed no statistically significant differences

P > 0.05). DUR was positively correlated with NP (rs = (Mann-Whitney U test, P > 0.05). Generally, 0.7, P < 0.05) while negatively with water temperature P. kessleri sounds were longer in DUR (Mann-Whit-

(rs = −0.6, P < 0.05). NP was also negatively correlated ney U test: N1 =7,N2 =8,U=0.0,P <0.05)andwith with water temperature (rs = −0.7, P < 0.05). Lastly, PRR higher NP (Mann-Whitney U test: N1 =7,N2 =8,U= was positively correlated with PF (rs =0.7,P < 0.05). Of 0.0, P < 0.05) and PF (Mann-Whitney U test: N1 =7, the six acoustic properties of tonal segments, SR, DUR, N2 =8,U = 11.0, P < 0.05; Fig. 3). Although not sig- PF and FM were more variable between- than within- nificant, P. kessleri vocalizations exhibited greater PRR individuals, as all CVb/CVw ratios >1.0, indicating a mean values compared to N. melanostomus (Fig. 3), higher level of inter-individual variability (Table 3). which in turn, had greater FM values than P. kessleri 732 Environ Biol Fish (2019) 102:727–739

Fig. 1 Temporal and spectral details of two consecutive vocali- spectrum (left) (window type: Hamming; FFT: 512; frame: 100; zations produced during reproductive interactions by resident bandwidth: 10 Hz; resolution: 8 Hz; overlap: 93.75%; filter band: Neogobius melanostomus (male #2, 144 mm LT). Above: oscillo- 50–500 Hz). Arrow on the power spectrum refers to the peak grams (left and right): below: spectrogram (right) and power frequency of produced sounds sequence

(Mann-Whitney U test: N1 =7,N2 =8,U=5.0,P <0.05). Discussion In addition, SR was also higher in N. melanostomus and differed significantly between the two species (Mann- The present study showed that the two investigated

Whitney U test: N1 =7,N2 =8,U=0.0, P < 0.05; Fig. Ponto-Caspian species previously placed in the same 3), highlighting once again the short duration of genus Neogobius sensu (Berg, 1949), Neogobius N. melanostomus sounds. melanostomus and Ponticola kessleri, produced distinct

Table 2 Descriptive statistics of the tonal sounds recorded from round goby Neogobius melanostomus and bighead goby Ponticla kessleri males

Acoustic variable Neogobius melanostomus Ponticola kessleri

x ± SD Range n x ± SD Range n P value

Sound rate (sounds per min.) 10.1 ± 4.4 1–46 7 2.5 ± 0.8 1–58<0.05 Duration (ms) 114.8 ± 10.4 50.5–217.1 7 457.9 ± 54.5 189.1–879.8 8 <0.05 Number of pulses 9.2 ± 0.4 5–15 7 44.9 ± 4.5 17–72 8 <0.05 Pulse repetition rate (Hz) 87.2 ± 5.9 55.9–124.8 7 99.9 ± 5.6 64.2–128.9 8 > 0.05 Peak frequency (Hz) 86 ± 6.5 52–128 7 105 ± 11.2 81–209.9 8 <0.05 Frequency modulation (Hz) 3.5 ± 13.5 −49.1 – (+56.3) 7 −9.2 ± 6.9 −39.2 – (+38.3) 8 <0.05

Individual means of the acoustic features of both species were compared with the Mann–Whitney U test P values refer to the results of the non-parametric Mann-Whitney U test, n-number of individuals acoustically tested. Sounds from N. melanostomus produced within reproductive and aggressive context were first compared against each other to test for differences, and since there were none based on all acoustic properties (Mann-Whitney U test, P > 0.05), they were pooled and compared with P. kessleri vocalizations Bold values indicate the sound properties that differed significantly between two investigated species except for PRR, emphasizing the acoustic differences in produced vocalizations Environ Biol Fish (2019) 102:727–739 733

Table 3 Within male variability (CVw) and between-male variability (CVb) for the six acoustic variables analysed for seven Neogobius melanostomus males and eight Ponticola kessleri males

Acoustic variable Neogobius melanostomus Ponticola kessleri

CVw CVb CVb/CVw CVw CVb CVb/CVw

Sound rate (sounds per min) 0.32 0.33 1.05 0.29 0.36 1.23 Duration (ms) 0.17 0.17 1.03 0.23 0.23 1.01 Number of pulses 0.14 0.15 1.01 0.21 0.20 0.97 Pulse repetition rate (Hz) 0.11 0.11 0.95 0.09 0.09 0.99 Peak frequency (Hz) 0.16 0.15 0.95 0.11 0.11 1.06 Frequency modulation (Hz) −1.13 3.80 −3.34 −1.07 −1.25 1.17

From five to ten sounds were analysed for each male of both species. Since no significant difference was found between aggressive and reproductive sounds produced by N. melanostomus (Man-Whitney U test; P > 0.05), the data were pooled and analysed together for the overall mean and standard deviation, and subsequently for CV coefficients and CVb/CVw ratio Bold features represent the acoustic variables which are more variable among- than within-males (CVb/CVw > 1), indicating that these features could potentially discriminate individuals vocalizations during intraspecific aggressive and repro- 1.0) could potentially discriminate males and/or be used ductive interactions, showing well distinguishable by females for mate assessment. In this study, there was species-specific sound structures. The Neogobius struc- strong stereotypy in the tonal sounds produced by both ture is characterised by short, low-frequency tonal N. melanostomus and P. kessleri, since certain temporal sounds, while the Ponticola sounds appear as a broad- acoustic features (mainly DUR and SR) showed larger band, frequency modulated longer call. According to between- than within-individual variation, and therefore Amorim et al. (2013), stereotyped features (CVb/CVw > presented a potential role for acoustic communication

Fig. 2 Temporal and spectral details of two consecutive vocali- 10 Hz; resolution: 8 Hz; overlap: 93.75%; filter band: 50–500 Hz). zations produced during aggressive interactions by resident Arrow on the power spectrum refers to the peak frequency of Ponticola kessleri (male #6, 171 mm LT). Above: oscillograms produced sounds sequence. Grey area on the spectrogram indi- (left and right): below: spectrogram (right) and power spectrum cates low-frequincy ambient noise (50–100 Hz) captured with the (left) (window type: Hamming; FFT: 512; frame: 100; bandwidth: hydrphone during aggressive trials 734 Environ Biol Fish (2019) 102:727–739

Fig. 3 Comparison between Neogobius melanostomus and (Mann-Whitney U test, *P < 0.05; **P < 0.01) Medians (point) Ponticola kessleri for sound rate, sound duration, cycle repetition and percentiles (whisker value) are depicted rate and peak frequency. * and ** indicate significant difference and individual discrimination. At the interspecific level, tested congeneric species, Neogobius fluviatilis, pro- communicative vocalizations differed markedly accord- duces the same sound type but with less repetition ing to acoustic parameters and sound structures, indicat- within the same time frame (Horvatić et al. 2015), again ing that each species possess fine-tuned temporal and confirming that congeneric species can be differentiated spectral characteristics enabling them to be differentiat- according to acoustic signals. In P. kessleri,soundsare ed between conspecifics. In this sense, N. melanostomus longer and scarcer, and this study indicates their infor- calls can be recognized as short tonal sounds organised mation content is contained within the frequency mod- in long series and intensively repeated (high sound rate), ulation, which also proved to be stereotyped, rather than while P. kessleri produces longer calls, which are down- in a series of repeated calls. It is worth mentioning that ward frequency modulated and rarely produced in long the sound emission rate could be a reliable indicator of trains, more frequently singly (low sound rate). The male size, condition and motivation (Bass and higher sound rate of N. melanostomus sounds seems to McKibben 2003;Amorimetal.2013), while Mitchell indicate that these shorter sounds could be produced et al. (2008) proposed that calling rate may be limited by more easily than longer frequency modulated calls, sonic muscle fatigue. In addition, long duration tonal without great energy consumption. However, future sounds are expected to generate fatigue more quickly, as studies dealing with the metabolic cost of sound pro- supported in the boatwhistles of Gulf toadfish, Opsanus duction are needed to clarify this hypothesis. In addition, beta (Thorson and Fine 2002). While physiological we suggest that sound rate, showing a stereotyped na- studies remain to be conducted, in this study, ture, could be an informative acoustic trait for N. melanostomus emitted calls during both aggressive N. melanostomus, since the only other acoustically and reproductive contexts, which could partially explain Environ Biol Fish (2019) 102:727–739 735

Table 4 Sound structure (oscillogram-above; spectrogram-below), systematic position, phylogenetic relationship and distribution of soniferous Ponto-Caspian and some Mediterranean gobies Species Systematic Distribution Sound Acoustic References position and structure properties phylogeny Padogobius Gobiidaea; Gobius Northern Adriatic Long Lugli et al. (1997), bonelli – lineageb; basin: (Po River duration, Lugli et al. (2003), (Bonaparte, proposed Ponto- basin, Italy) and downward and Malavasi et al. 1846) Caspian originc (Istrian Peninsula; frequency- (2008) (acoustics); Zrmanja and Krka modulated Penzo et al. (1998), River basins, calls repeated Huyse et al. (2004) Croatia); with high (genetics) introduced into intensity rivers of the west coast of Italy Gobius Gobiidaea; Gobius Eastern Atlantic Long Malavasi et al. paganellus – lineageb; Ocean, duration, (2008) and Linnaeus, Mediterranean and Mediterranean upward Parmentier et al. 1758 North-Eastern Sea and Black frequency- (2013) (acoustics); Atlantic gobiesc Sea modulated Penzo et al. (1998), calls repeated Huyse et al. (2004) with high (genetics) intensity Ponticola Gobiidaea; Gobius Black Sea basin; Long This study kessleri – lineageb; Ponto- invasive in duration, (acoustics); Neilson (Günther, Caspian gobiesc; western and downward & Stepien (2009), 1861) Ponticolinid central Europe frequency- Medvedev et al. modulated (2013) (genetics) calls repeated with low intensity Proterorhinus Gobiidaea; Gobius Black Sea basin; Not available Medium long, Ladich & Kratochvil marmoratus – lineageb; Ponto- invasive in probably (1989) (acoustics); (Pallas, 1814) Caspian gobiesc; Danube basin repeated in Stepien & Tumeo Ponticolinid since the 1970s series of (2006), Neilson & moderate Stepien (2009) intensity (genetics) Neogobius Gobiidaea; Gobius Azov and Black Short Horvatić et al. (2015) fluvitilis – lineageb; Ponto- Sea basin; invasive duration, (acoustics); Neilson (Pallas, 1814) Caspian gobiesc; in western and upward & Stepien (2009), Neogobiinid central Europe frequency- Medvedev et al. modulated (2013), Šanda & calls repeated Kovačić & Šanda (2016) with moderate (genetics) intensity Neogobius Gobiidaea; Gobius Azov, Black and Short This study melanostomus – lineageb; Ponto- Caspian Sea basins; duration, (acoustics); Neilson (Pallas, 1814) Caspian gobiesc; invasive in eastern, upward & Stepien (2009), Neogobiinid western and central frequency- Medvedev et al. Europe and Great modulated (2013), Šanda & Lakes (North calls repeated Kovačić & Šanda (2016) America) with high (genetics) intensity Padogobius Gobiidaea; Gobius Central west Italy, Medium long, Lugli et al. (1996) nigricans – lineageb; rivers Arno, upward (acoustics); Penzo et (Canestrini, proposed Ponto- Ombrone, Tiber, frequency al. (1998), Huyse et 1867) Caspian originc Amaseno modulated al. (2004) (genetics) calls repeated with high intensity a According to Thacker (2011) b According to Agorreta et al. (2013). Ponto-Caspian gobies were deeply nested within the Gobius-lineage, together with other large Atlantic- Mediterranean gobies c According to Thacker and Roje (2011). Authors subdivided the European gobies into three groups: Ponto-Caspian gobies, Mediterranean and North-Eastern Atlantic gobies According to Neilson and Stepien (2009). Authors subdivided subfamily Benthophilinae on three well supported tribes: Ponticolini, dNeogobiini and Benthohpilini 736 Environ Biol Fish (2019) 102:727–739 the higher sound rate compared to P. kessleri sounds, lack the swim bladder (Lugli et al. 1996;Miller2004). since it is well know that sound rate increases when the In contrast, according to literature, many other European emitter is courting or facing a potential female intruder gobies sharing a close affinity with P. nigricans (i.e., (Mann and Lobel 1995;Amorim2006). Regardless of Padogobius bonelli and Gobius paganellus)have inter- or intrasexual interactions, the general pattern of retained the swim bladder through all life stages, which interspecific differences remains the same, taking into poses an interesting question about the evolutionary consideration the sound complexity and sound structure. pattern of swim bladder loss or retention in this group Our results are consistent with Malavasi et al. (2008), of species, and possible relationships with sound struc- who suggested that the temporal patterning of sounds, ture and emission mechanisms. In rocky goby i.e. DUR and PRR, are the acoustical properties respon- (G. paganellus), cranial-pectoral muscles (m. levator sible for discriminating between the Mediterranean pectoralis) contractions were responsible for the pro- gobies. Acoustic differences from this study seem also duction of long, frequency modulated tonal sounds to be consistent with previous phylogenetic studies, (Parmentier et al. 2013)similartoP. kessleri vocaliza- which separated species from the former Neogobius tions. Further investigations on sound emission mecha- genus into additional tribes, namely Neogobini nisms should clarify whether this cranio-pectoral system (Neogobius melanostomus, N. fluviatilis and is common to all goby acoustic systems or, alternatively, N. caspius), Benthophilini (tadpole gobies) and whether different mechanisms are used by different Ponticolini (genera Mesogobius, Proterorhinus, Babka, groups of species with different sound structures, such and Ponticola) (Neilson and Stepien, 2009;Medvedev as the Bneogobini^ group. These observations raise the et al. 2013). In addition, our results indicate that there is question about the systematic position of P. nigricans, a strong similarity of N. melanostomus sounds with and its relative P. bonelli, another Italian goby endemism, those documented from another vocal con-generic spe- since many authors have previously mentioned that this cies, i.e. the monkey goby N. fluvitalis (Horvatić et al. species corresponds to different genera and that they 2015). From a comparative perspective, the two could be more closely related to Ponto-Caspian lineage Neogobius species, sharing a close genetic and morpho- than to European gobies (Lugli et al. 1996; Penzo et al. logical relationship (Miller 2003; Neilson and Stepien 1998; Huyse et al. 2004; Miller 2004; Kottelat and 2009;Medvedevetal.2013;Kovačić and Šanda 2016), Freyhof 2007). Observed acoustic similarities parallels were able to produce same sound type, with a strong the molecular affinities revealed by recent phylogenetic overlap in the general structure and acoustic behaviour. studies, which suggested that Ponto-Caspian and Medi- While this resemblance in acoustic features is not so terranean gobies share close phylogenetic relationships surprising for con-generic species, more striking is the (Thacker and Roje 2011; Agorreta et al. 2013;Thacker similarity of acoustic signals for the two Neogobius 2015). In Proterorhinus marmoratus, another soniferous species and another vocal Mediterranean species, the Ponto-Caspian species, the vocal repertoire is strictly Arno goby, Padogobius nigricans. P. nigricans is an tonal and both sexes emit short duration (around endemic Italian freshwater fish, geographically well 250 ms), low frequency sounds, (Ladich and Kratochvil separated from rest of the Ponto-Caspian basin (Miller 1989) which appear very similar in structure to vocaliza- 2004). The oscillogram and the power spectrum of tions described for Neogobius spp., with minor acoustic P. nigricans and Neogobius spp. sounds mostly overlap, differences. According to Neilson and Stepien (2009) with mean acoustic values varying slightly between and Thacker and Roje (2011), P. marmoratus is posi- species (Lugli et al. 1996, 1997). According to morpho- tioned as a basal member within the tribe Ponticolini, and logical studies, a close relationship between the genus regarding its acoustic traits, it could represent a transi- Padogobius and the Ponto-Caspian clade was previous- tional taxon between the former tribe and Neogobiini. On ly suggested on the basis of morphological traits the other hand, P. kessleri produces vocalizations that on (Bianco and Miller 1990; Simonovic et al., 1996; oscillogram mostly resemble the sounds emitted by two Miller 2003, 2004), indicating that Ponto-Caspian group vocal Mediterranean species, G. paganellus and is synapomorphic with the Italian freshwater P. bonelli (tonal signals), with minor differences in mean Padogobius in possessing a higher number of vertebrae, acoustic values (Lugli et al. 1995, 1997; Malavasi et al. but also having a reduced or atrophied swim bladder. 2008; Parmentier et al. 2013). Tonal-like sounds pro- Interestingly, both Neogobius species and P. nigricans duced by these Mediterranean species are characterized Environ Biol Fish (2019) 102:727–739 737 by a high pulse repetition rate (around 100 Hz) of long (present study; Ladich and Kratochvil 1989; Horvatić duration (> 300 ms) (Zeyl et al. 2016), and overlapping in et al. 2015). In conclusion, it is important to note that values with the P. k ess le ri vocalizations. Penzo et al. different soniferous European gobies produce similar (1998) and Huyse et al. (2004) documented the close types of vocalizations, exemplifying the continuum from affinity between G. paganellus and P. nigricans,empha- pulsatile to tonal sounds, though each species quantita- sizing that the genus Padogobius could be paraphyletic. tively modifies its vocal repertoire, allowing specific iden- Considering the present observations, G. paganellus tification. The present study suggests that the two inves- acoustically could be more closely related to P. bonelli tigated species belonging to the subfamily than to P. nigricans, again supporting the complex ex- Benthophilinae, Neogobius melanostomus and Ponticola planation of this phylogenetic relationship. From the kessleri, were able to produce frequency modulated tonal results of this study and the existing literature on vocal sounds, but each taxon modified their acoustical features European gobies, two groups of species producing tonal- and developed vocalizations differently according to their like sounds can be recognized (Table 4). Ponto-Caspian phylogenetic history. From the phylogenetic context, vo- N. fluviatilis, N. melanostomus, P. marmoratus and the calizations produced by gobies from Benthophilinae sub- Arno goby P. nigricans form the first group, while family diverge according to existing molecular studies P. kessleri, P. bonelli and G. paganellus constitute the (Neilson and Stepien 2009; Thacker and Roje 2011; second group of soniferous gobies, differentiated accord- Medvedev et al. 2013) and proposed tribes can be well ing to their signal properties. By exploring the freshwater recognized according to sound structure and acoustic lifestyle history of Mediterranean gobies, Penzo et al. traits. Species from the Neogobiini tribe sensu Neilson (1998) proposed that Padogobius spp. and and Stepien (2009) are characterized by producing inten- G. paganellus originated from the same ancestor, while sive, repeated, short, tonal sounds, while the only Breal^ the former genus retained its freshwater distribution, fur- Ponticolini species acoustically tested, P. kessleri,produce ther subdivided during the Messinian salinity crisis (MSC; longer, downward frequency modulated tonal calls. These 5.9–5.3 Mya), while the latter secondarily recolonized scenarios emphasize that acoustic properties should not be marine habitats and nowadays inhabits the Eastern Atlan- neglected when reconstructing phylogenetic relationships, tic, Mediterranean and Ponto-Caspian area (Kovačić and especially for the Ponto-Caspian group, which have Patzner 2011). According to this scenario, the Ponto- proved to contain vocal taxa with well-developed sound Caspian ancestor migrated through the Pannonian channel production abilities. The recorded vocal repertoire present to the Proto-Mediterranean (24–15 Mya) and consequent- a baseline for the future comparative acoustic studies ly adapted to new ecological niches during the later stage within the subfamily Benthophilinae, aiming to gain ad- of MSC known as BLago Mare^ (5.5 Mya), during which ditional information on the evolution of acoustic commu- habitats gradually alternated from freshwater and brackish nication within Ponto-Caspian gobies and highlighting (< 5.3 Mya) to exclusively marine (> 5.3 Mya). Although their importance in reconstructing phylogenetic speculative, the relationship between G. paganellus and relationships. the Ponto-Caspian gobies remains unknown and future phylogeographical studies are needed to clarify this hy- Acknowledgements Authors are grateful to Linda Zanella for pothesis and the patterns of goby diversification through- proofreading the manuscript and two anonymous referees for their out the Mediterranean basin. In addition, this grouping comments that improved the overall quality of the manuscript. All emphasizes that there is a certain degree of variability methods and procedures were approved by the Ethics Committee of Biology Division (Faculty of Science, approval no: 251-58- between species within the same genus (i.e. Padogobius) 10617-18-14). and that detailed phylogenetic analysis is urgently needed to fully resolve their systematic position. Based on our results, the presence of sound production within the benthophilin gobies showed that acoustic communication References is a synapomorphy for this subfamily and that acoustic signals play a crucial role in their life history. Also, tonal č ć sounds seems to have a deeper origin within Agorreta A, San Mauro D, Schliewen U, Van Tassell JL, Kova i M, Zardoya R, Ruber L (2013) Molecular phylogenetics of Benthophilinae subfamily, since all acoustically investi- Gobioidei and phylogenetic placement of European gobies. gated species were able to produce this sound type Mol Phylogenet Evol 69:619–633 738 Environ Biol Fish (2019) 102:727–739

Akamatsu A, Okumura T, Novarini N, Yan HY (2002) Empirical Kottelat M, Freyhof J (2007) Handbook of European freshwater refinements applicable to the recording of fish sounds in fishes. Kouelat, Comol, Switzerland and Freyhof. Berlin, small tanks. J Acoust Soc Am 112:3073–3082 Gennany Amorim MCP (2006) Diversity of sound production in fish. In: Kovačić M, Patzner RA (2011) North-Eastern Atlantic and Ladich F, Collin SP, Moller P, Kapoor BG (eds) Mediterranean gobius. In: Patzner RA, Van Tassell JL, Communication in fishes, vol I. Science Publishers, Kovačić M, Kapoor BG (eds) The biology of gobies. Enfield, pp 71–105 Science Publishers, CRC Press, Taylor & Francis Group, Amorim MCP, Neves ASM (2007) Acoustic signalling during New York, pp 177–206 courtship in the painted goby, Pomatoschistus pictus.JMar Kovačić M, Šanda R (2016) A new species of Gobius Biol Assoc UK 87:1017–1023 (Perciformes: Gobiidae) from the Mediterranean Sea and the Amorim MCP, Simões JM, Almada VC, Fonseca PJ (2011) redescription of Gobius bucchichi. J Fish Biol 88:1104–1124 Stereotypy and variation of the mating call in the Lusitanian Ladich F, Kratochvil H (1989) Sound production in the marmoreal toadfish, Halobatrachus didactylus. Behavioral Ecology and goby Proterorhinus marmoratus (Pallas) (Gobiidae: Sociobiology 65(4):707–716 Teleostei). Zool Jahrb Abt allg Zool Physiol Tiere 93:501–504 Amorim MCP, Pedroso SS, Bolgan M, Jordão JM, Caiano M, Laiolo P, Rolando A (2003) The evolution of vocalisations in the Fonseca PJ (2013) Painted gobies sing their quality out loud: genus Corvus: effects of phylogeny, morphology and habitat. acoustic rather than visual signals advertise male quality and Evol Ecol 17:111–123 contribute to mating success. Funct Ecol 27:289–298. Lobel PS (1998) Possible species specific courtship sounds by two https://doi.org/10.1111/13652435.12032 sympatric cichlids fishes in Lake Malati, Africa. Environ Biol – Bass AH, Mckibben JR (2003) Neural mechanisms and behaviors Fish 52:443 452 for acoustic communication in teleost fish. Neurobiology Lugli M (2010) Sounds of shallow water fishes pitch within the 605:1–26 quiet window of the habitat ambient noise. J Comp Physiol A – Berg LS (1949) Ryby presnykh vod SSSR i sopredel’nykh stran. 196:439 451. https://doi.org/10.1007/s00359-010-0528-2 Ch. 3 (freshwater fishes of the USSR and adjacent countries. Lugli M, Torricelli P (1999) Prespawning sound production in – Part 3, Moscow: Akad. Nauk SSSR mediterranean sand gobies. J Fish Biol 54:691 694 Bianco PG, Miller PJ (1990) Yugoslavian and other records of the Lugli M, Pavan G, Torricelli P, Bobbio L (1995) Spawning Italian freshwater goby Padogobius martensii, and a charac- vocalisations in male freshwater gobiids (Pisces, Gobiidae). – ter polarization in gobioid fishes. J Nat Hist 24:1289–1302. Environ Biol Fish 43:219 231 https://doi.org/10.1080/00222939000770761 Lugli M, Torricelli P, Pavan G, Miller PJ (1996) Breeding sounds of male Padogobius nigricans with suggestions for further Bradbury JW, Vehrencamp SL (1998) Principles of animal com- evolutionary study of vocal behaviour in gobioid fishes. J munication. Sinauer Associates, Sunderland Fish Biol 49:648–657 Cap H, Deleporte P, Joachim J, Reby D (2008) Male vocal Lugli M, Torricelli P, Pavan G, Mainardi D (1997) Sound produc- behavior and phylogeny in deer. Cladistics 24:917–931 tion during courtship and spawning among freshwater Christie PJ, Mennill DJ, Ratcliffe LM (2004) Chickadee song gobiids (Pisces, Gobiidae). Mar Fresh Behav Physiol 29: structure is individually distinctive over long broadcast dis- 109–126 tances. Behav 141:101–124 Lugli M, Yan HY, Fine ML (2003) Acoustic communication in Cocroft RB, Ryan MJ (1995) Patterns of advertisement call evo- two freshwater gobies: the relationship between ambient – lution in toads and chorus frogs. Animal Behav 49:283 303 noise, hearing thresholds and sound spectrum. J Comp Colleye O, Vandewalle P, Lanterbecq D, Lecchini D, Parmentier E Physiol 189:309–320 (2011) Interspecific variation of calls in clownfishes: degree of Malavasi S, Torricelli P, Lugli M, Pravoni F, Mainardi D (2003) similarity in closely related species. BMC Evol Biol 11:365 Male courtship sounds in a teleost with alternative reproduc- Crawford JD (1997) Hearing and acoustic communication in tive tactics, the grass goby, Zosterisessor ophiocephalus. – mormyrid electric fishes. Mar Freshw Behav Physiol 29:65 86 Environ Biol Fish 66:231–236. https://doi.org/10.1023 Gingras B, Mohandesan E, Boko D, Fitch WT (2013) /A:1023923403180 Phylogenetic signal in the acoustic parameters of the adver- Malavasi S, Collatuzzo S, Torricelli P (2008) Interspecific varia- tisement calls of four clades of anurans. BMC Evol Biol 13: tion of acoustic signals in Mediterranean gobies 134. https://doi.org/10.1186/1471-2148-13-134 (Perciformes, Gobiidae): comparative analysis and evolu- Günther AC (1861) Catalogue of Fishes in the British Museum 3: tionary outlook. Biol J Linn Soc 93:763–778. https://doi. 409–467 org/10.1111/j.1095-8312.2008.00947.x Hamao S, Sugit N, Nishiumi I (2015) Geographic variation in bird Malavasi S, Valerio C, Torricelli P (2009) Courtship sounds and songs: examination of the effects of sympatric related species associated behaviours in the Canestrini’sgoby on the acoustic structure of songs. Acta Ethol 19:1–10. Pomatoschistus canestrinii. J Fish Biol 75:1883–1887 https://doi.org/10.1007/s10211-015-0228-6 Mann DA, Lobel PS (1995) Passive acoustic detection of sounds Horvatić S, Cavraro F, Zanella D, Malavasi S (2015) Sound produced by the damselfish Dascyllus albisella production in the Ponto-Caspian goby Neogobius fluviatilis (Pomacentridae). Bioacoustics 6:199–213 and acoustic affinities within the Gobius lineage: implica- Medvedev DA, Sorokin PA, Vasil’ev VP, Chernova NV, Vasil’eva tions for phylogeny. Biol J Linn Soc 117:564–573 ED (2013) Reconstruction of phylogenetic relations of Huyse T, Van Houdt J, Volckaert FAM (2004) Paleoclimatic Ponto-Caspian gobies (Gobiidae, Perciformes) based on mi- history and vicariant speciation in the ‘sand goby’ group tochondrial genome variation and some problems of their (Gobiidae, Teleostei). Mol Phylogenet Evol 32:324–336 taxonomy. J Ichthyol 53:702–712 Environ Biol Fish (2019) 102:727–739 739

Mélotte G, Vigouroux R, Michel C, Parmentier E (2016) Rollo A, Andraso G, Janssen J, Higgs D (2007) Attraction and Interspecific variation of warning calls in piranhas: a com- localization of round goby (Neogobius melanostomus)to parative analysis. Sci Rep 6:36127. https://doi.org/10.1038 conspecific calls. Behaviour 144:1–21 /srep36127 Sebastianutto L, Picciulin M, Costantini M, Rocca M, Ferrero E Miller PJ (1984) The tokology of gobioid fishes. In: Potts GW, (2008) Four types of sounds from one winner: vocalizations Wotton RJ (eds) Fish reproduction: strategies and tactics. during territorial behaviour in the red-mouthed goby Gobius Academic Press, London, pp 119–153 cruentatus (Pisces, Gobiidae). Acta Ethol 11:115–121 Miller PJ (2003) The freshwater fishes of Europe. Mugilidae, Simonovic PD, Nikolić VP, Skora KE (1996) Vertebral number in Atherinidae, Atherinopsidae, Blenniidae, Odontobutidae, Ponto-Caspian gobies: phylogenetic relevance. J Fish Biol Gobiidae 1, Wiebelsheim: AULAVerlag, vol. 8/I, pp 157-345 49:1027–1029 Miller PJ (2004) The freshwater fishes of Europe. Gobiidae 2, Spanier E (1979) Aspects of species recognition by sound in four Wiebelsheim: AULA Verlag, vol. 8/II, pp 33-93 species of damselfishes, genus Eupomacentrus (Pisces: Mitchell S, Poland J, Fine ML (2008) Does muscle fatigue limit Pomacentridae). Z Tierpsychol 51:301– 316. https://doi. advertisement calling in the oyster toadfish Opsanus tau? org/10.1111/j.1439-0310.1979.tb00691.x Animal Behav 76:1011–1016 Thacker CE (2011) Systematics of Gobiidae. In: Patzner RA, Van Myrberg AA, Lugli M (2006) Reproductive behavior and acous- Tassell JL, Kovačić M, Kapoor BG (eds) The Biology of tical interations. In: Ladich F, Collin SP, Moller P, Kaapor BG Gobies. Enfield, NH: Science Publishers, p 129–136 (eds) Communication in fishes, vol 1. Science Publishers, Thacker CE (2015) Biogeography of goby lineages (Gobiiformes: – Enfield, pp 146 176 Gobioidei): origin, invasions, and extinction through the Myrberg AA, Spanier E, Ha SJ (1978) Temporal patterning in Cenozoic. J Biogeogr 42:1615–1625. https://doi. acoustical communication. Contrasts in Behavior, Wiley, org/10.1111/jbi.12545 New York Thacker CE, Roje DM (2011) Phylogeny of Gobiidae and identi- Neilson ME, Stepien CA (2009) Escape from the Ponto Caspian: fication of gobiid lineages. Syst Biodivers 9:229–347 evolution and biogeography of an endemic goby species Thorson RF, Fine ML (2002) Crepuscular changes in emission rate flock (Benthophilinae: Gobiidae: Teleostei). Mol and parameters of the boatwhistle advertisement call of the – Phylogenet Evol 52:84 102 gulf toadfish, Opsanus beta. Environ Biol Fish 63:321–331. Parmentier E, Kever L, Boyle K, Corbisier Y, Sawelew L, https://doi.org/10.1023/A:1014334425821 Malavasi S (2013) Sound production mechanism in Gobius – Torricelli P, Parmigiani S, Lugli M, Gandolfi G (1988) Intermale paganellus (Gobiidae). J Exp Biol 216:3189 3199 aggression in Padogobius martensi (günther) (pisces Penzo E, Gandolfi G, Bargelloni L, Colombo L, Patarnello T Gobiidae): effect of size and prior residence. Monit Zoolog (1998) Messinian salinity crisis and the origin of freshwater Ital – Ital J Zool 22:121–131. https://doi.org/10.1080 lifestyle in western Mediterranean gobies. Mol Biol Evol 15: /00269786.1988.10736547 1472–1480 Torricelli P, Lugli M, Pavan G (1990) Analysis of sound produced Remage-Healey L, Bass AH (2006) A rapid, neuromodulatory by male Padogobius martensi (Pisces: Gobiidae) and factors role for steroid hormones in the control of reproductive affecting their structural properties. Bioacoustics 2:261–175 behavior. Brain Res 1126:27–35 Zeyl JN, Malavasi S, Holt DE, Noel P, Lugli M, Johnston CE Rice AN, Bass AH (2009) Novel vocal repertoire and paired (2016) Convergent aspects of acoustic communication in swimbladders of the three-spined toadfish, Batrachomoeus darters, sculpins and gobies. In: Sisneros AJ (ed) Fish hearing trispinosus: insights into the diversity of the Batrachoididae. J and bioacoustics: an anthology in honor of Arthur N. Popper Exp Biol 212:1377–1391 and Richard R. Fay. Springer International Publishing, Robillard T, Desutter-Grandcolas L (2004) High frequency calling Cham, pp 93–120 in Eneopterinae crickets (Orthoptera, Grylloidea, Eneopteridae): adaptive radiation revealed by phylogenetic analysis. Biol J Linn Soc 83:577–584 Publisher’snote Springer Nature remains neutral with regard to Robillard T, Legendre F, Desutter-Grandcolas L, Grandcolas P jurisdictional claims in published maps and institutional (2006) Phylogenetic analysis and alignment of behavioral affiliations. sequences by direct optimization. Cladistics 22:602–633 Journal of Zoology. Print ISSN 0952-8369

Acoustic communication during reproduction in the basal gobioid Amur sleeper and the putative sound production mechanism S. Horvatic1, S. Malavasi2,E.Parmentier3,Z.Marci c1,I.Buj1,P.Mustaﬁc1,M.Caleta 4, M. Smederevac-Lalic5, S. Skoric5 & D. Zanella1

1 Department of Zoology, Faculty of Science, University of Zagreb, Zagreb, Croatia 2 Department Environmental Sciences, Informatics and Statistics, Ca’ Foscari University of Venice, Venezia Mestre, Italy 3 Laboratoire de Morphologie Fonctionnelle et Evolutive, AFFISH-RC, Institut de Chimie – B6C, UniversitedeLi ege, Liege, Belgium 4 Faculty of Teacher Education, University of Zagreb, Zagreb, Croatia 5 Institute for Multidisciplinary Research, University of Belgrade, Belgrade, Serbia

Keywords Abstract Perccottus glenii; vocal fish; tonal sounds; sound production; acoustic behaviour; levator pectoralis Gobioids (Gobiiformes: Gobioidei) are a large group of vocal fishes with four muscle; acoustic communication; gobioids. sound types documented during aggressive or reproductive interactions in 23 species. Most attention has been dedicated to sound production in Gobiidae and Correspondence Gobionellidae, while acoustic communications in other phylogenetically distant Davor Zanella, Department of Zoology, Faculty of gobioid groups have been neglected. Odontobutidae, a basal family within the gob- Science, University of Zagreb, Rooseveltov trg 6, ioids, is a poorly studied fish assemblage, with sounds documented in only a single 10000, Zagreb, Croatia. Tel: +385(0)1 6189 709 species. The goal of this study was to record and describe the acoustic signals pro- Email: [email protected] duced by Perccottus glenii (Odontobutidae) under laboratory conditions, with particular focus on the reproductive phase (courtship and pre-spawning), and to Editor: Jean-Nicolas Volff provide insight into the anatomical basis of the sound emission mechanism. We recorded two acoustically different call types, thumps and tonal sounds. Thumps Received 3 January 2019; revised 25 June 2019; were low-frequency sounds (~95 Hz) with an irregular waveform, produced by accepted 1 July 2019 males during both the courtship and pre-spawning phases. Thumps were frequently organized in long trains, a thump burst, composed from approximately five thumps doi:10.1111/jzo.12719 and lasting over 10 s. Tonal sounds were short vocalizations (~90 ms) produced only during courtship interactions, characterized by a sinusoidal oscillogram and a single frequency peak (~120 Hz). Additionally, anatomical examination focusing on the pectoral girdle identified the muscles that could be responsible for sound emission. The levator pectoralis muscle, originating on the neurocranium and attaching to the cleithral bone, is separated into three bundles: a pars lateralis superficialis, a pars lateralis profundus and a pars medialis. These results expand the knowledge about gobioid vocal behaviour and underline the importance of acoustic communication within this group of fish. Odontobutidae is a sister group to rest of the gobioids, and therefore, our results have significant impact for future comparative studies dealing with sound production.

Introduction subdivided into the courtship, pre-spawning and spawning phases (Lugli et al., 1997; Myrberg & Lugli, 2006; Amorim & Although numerous ﬁshes can produce different sounds, their Neves, 2007; Malavasi, Valerio & Torricelli, 2009; Blom production for social communication is documented in a et al., 2016; Zeyl et al., 2016). Courtship behaviour usually restricted number of species from around 110 families (Par- commences from the moment the female enters the male terri- mentier & Fine, 2016). Gobioid ﬁshes (phylogeny according to tory, while the pre-spawning phase is characterized with the Thacker, 2009) form one of the most investigated Teleost ripe female entering the male’s nest. When the female inverts groups in terms of vocal repertoire. Within this group, acoustic in the nest to release eggs onto the ceiling, spawning has communication has been documented in 23 species (Horvatić started. et al., 2015; Zeyl et al., 2016). In general, gobioid sound Four major sound types can be recognized in gobioids: emission occurs during agonistic and reproductive contexts. pulsed sounds consisting of pulse trains repeated at a slow Acoustically, the reproductive sequence can be further rate, tonal sounds characterized by faster pulse repetition rates

Journal of Zoology 309 (2019) 269–279 ª 2019 The Zoological Society of London 269 Perccottus glenii sounds and vocal mechanism S. Horvatic et al. and a sinusoidal-like waveform, and complex sounds that are a acoustic structure within the gobioid fishes. The spawning combination of the two (Lugli et al., 1997; Zeyl et al., 2016). habits of P. glenii are similar to the general goby pattern, and Lastly, thump sounds are short low-frequency pulsed signals during reproduction, which starts in May or July, P. glenii with sound energy under 1 kHz (Amorim & Neves, 2007; males occupy and defend nests that are used as spawning sites Malavasi, Collatuzzo & Torricelli, 2008; Zeyl et al., 2016). and exhibit territorial behaviour (Miller, 1986; Kottelat & Frey- Briefly, in the most investigated species, pulse trains dominate hof, 2007). during both agonistic and reproductive (pre-spawning) interac- The aim of the present study was to assess the patterns of tions, while thumps and tonal sounds are involved in courtship acoustic communication in P. glenii during courtship while activities in many gobioids (Ladich & Kratochvil, 1989; Lugli also providing insight into the anatomical basis of the sound et al., 1996; Lindstrom€ & Lugli, 2000; Amorim & Neves, emission mechanism. More specifically, the goals were to: (1) 2007, 2008; Malavasi et al., 2008, 2009; Horvatić et al., 2015; document sound production during reproduction in the male Zeyl et al., 2016). Complex sounds, reported in only three P. glenii and quantitatively describe the sound structure; (2) gobioids to date, were recorded during aggressive circum- explore the association between reproductive behaviours and stances (Gobius cruentatus and Periophthalmodon septemra- sound production; and (3) investigate the morphological com- diatus) or as a part of spawning acoustic repertoire in ponents of the pectoral girdle and associated muscles possibly Padogobius bonelli (Lugli et al., 1995, 1997; Sebastianutto related to sound production in the odontobutid P. glenii. et al., 2008; Polgar et al., 2011; Zeyl et al., 2016). Recently, molecular studies of the relationships between the Materials and methods gobioids strongly support that Odontobutidae is a sister group of all other lineages (Thacker, 2009; Agorreta et al., 2013). The study was carried out from April until the end of June This group contains 22 species, with acoustic communication 2018, which corresponds with the breeding season of P. glenii. studied in just one species: Odontobutis obscura (Takemura, Fish were caught in October 2017 from a well-vegetated tribu- 1984). The acoustic feature of its recorded calls resembles the tary to the main Danube River channel, near Veliko Gradiste pulsatile sounds previously documented in many other gobies (Serbia). Individuals (25 individuals; 10 males and 15 females) (see Zeyl et al., 2016). These results support the claim that were collected using electrofishing and transported to the labo- sound production based on pulsed repertoires could be a ratory where they were sexed according to urogenital papillae synapomorphic trait within the gobioids (Malavasi et al., 2008; (conical in males, wider and fimbriose in females; Miller, Zeyl et al., 2016). However, additional acoustic research is 2003) and separated by sex into two stock glass tanks (each required to understand the ancestral state of sound communica- 120 L). Six males exhibiting pronounced breeding behaviour tion in gobioid fishes. with obvious dimorphic traits (swollen nape and dark col- The diversity of sonic mechanisms in fishes is so great that oration) were used as residents in acoustic tests and placed it has not been possible to group them in appropriate cate- individually in 90-L glass tanks. Each tank was supplied with gories (Ladich & Fine, 2006). In gobioids, different mecha- a tile shelter to serve as a nest, aerators to ensure optimal oxy- nisms have been proposed; however, these hypotheses should gen levels and a gravel or fine sand substrate layer on the bot- be considered with caution as they have rarely been tested, tom. To avoid sound disturbance and minimize the and deeper multidisciplinary investigations are needed to fully surrounding noise during audio recordings, each tank was understand sound production in the gobioids. Parmentier et al. placed on foam-rubber shims. Photoperiod and water tempera- (2013, 2017a) experimentally tested the hypothesis of contrac- ture followed natural environmental reign, and in situations tion of the pectoral girdle muscles in two European gobies, when water temperature dropped below 15°C, water heater Gobius paganellus and Pomatoschistus pictus (Gobiidae and (ATMAN brand; AT 1008 type; temp. range: 20–34°C) was Gobionellidae, respectively). Multidisciplinary investigation used to maintain optimal water temperatures for reproduction (sound analysis, electromyography and electron microscopy), (Miller, 2003). Inside each tank, water temperature (range: together with morphological and high-speed videos, has sug- 15.7–20.8°C) was monitored with the thermometer (AQUA- gested that drumming sounds might be generated by the peri- TERRA, Garesnica, Croatia). odic contraction of the levator pectoralis muscle (Parmentier et al., 2013, 2017a). These findings proposed that the pectoral Experimental procedure girdle mechanism could be common to the gobies and may have possibly evolved in other gobioid groups as an exaptation Sounds were recorded from six males (size range: 84–96 mm (Parmentier, Diogo & Fine, 2017b), likely from locomotory LS), during the reproductive interactions (courtship and pre- movements. spawning phases). Experiments were set up according to the Perccottus glenii Dybowski, 1877 is a freshwater gobioid protocol provided by Horvatić et al. (2015) to record reproduc- fish, widespread in the lentic waters, ponds and marshes of tive (male–female) or aggressive interactions (male–male). We Europe and Central Asia (Kottelat & Freyhof, 2007; Freyhof, used an acoustic intruder test, during which the fish was left in 2011). It belongs to the family Odontobutidae, an early the observational tank for acclimatization (10–15 days) to branching lineage of the Gobiiformes, native to East Asia establish territoriality. An intruder (i.e. a female showing an (Kottelat & Freyhof, 2007; Thacker, 2009), and as such offers enlarged belly and nuptial coloration, or aggressive male) was a great opportunity to expand the knowledge on the vocal selected from the stock tank. The intruder fish (size range: 73– repertoire in Odontobutidae and to better clarify the ancestral

270 Journal of Zoology 309 (2019) 269–279 ª 2019 The Zoological Society of London S. Horvatic et al. Perccottus glenii sounds and vocal mechanism

89 mm LS) was placed in metal cage in front of the nest in the software allowed us to separate two different thump bursts the observational tank to elicit a behavioural or acoustic according to a time interval longer than 2-s, which separated response from the resident male. Each session lasted c. the last thump in one and the first thump in an additional 30 min. All pump devices, aerators and lights were switched burst. All thumps in the sequence with an interval shorter than off 10 min prior to recordings to minimize ambient noise. 2-s were considered one thump burst. The following acoustic properties were measured for thumps and thump bursts, respec- Sound collection tively: (1) thump rate (number of thumps emitted in 1 min; sound minÀ1); (2) thump duration (total length of the call, Sound production was recorded at various times during the measured in milliseconds); (3) number of cycles; (4) peak fre- day with an omnidirectional H2A-XLR hydrophone (Aquarian quency (obtained from the power spectrum function); (5) burst Audio & Scientific, Anacortes, WA, USA; sensitivity: duration (total length of the thump burst, measured in sec- À À180 dB re: 1 V l 1 Pa; frequency range Æ4 dB from 0.01 onds); (6) number of thumps in the burst; and (7) thump inter- to 100 kHz) positioned less than 10 cm from the nest entrance val (time interval within the burst measured from the end of according to Akamatsu et al. (2002). The hydrophone was one thump to the start of the following one, measured in sec- connected to an IRIG PRE preamplifier (Aquarian Audio & onds). For tonal sounds, we measured the following: (1) sound Scientific), and sounds were recorded using a ZOOM H4n por- rate (tonal sounds emitted in 1 min; sound minÀ1); (2) duration table digital audio recorder (wav 16/44.1 k mono; ZOOM, (ms); (3) number of cycles; (4) cycle repetition rate (Hz); (5) Tokyo, Japan), stored on the recorder memory card and moni- frequency modulation (after the sound has been divided into tored on headphones. three sections, frequency modulation was calculated as the difference between the final and initial cycle repetition rate and Behavioural recording expressed in Hz); and (6) peak frequency (Hz). The courtship and pre-spawning phases of reproductive beha- Statistical analysis viour were video-recorded and analysed. Courtship behaviour began when the females entered into the male territory, while The mean, standard deviation (x Æ SD) and range for each the pre-spawning phase was observed when the ripe female acoustic variable emitted by individual fish were calculated. entered the male’s nest. No sounds were produced during Sound features were examined using from 10 to 20 thumps, actual spawning. Video recordings of acoustic behaviour were 10 thump bursts and 5 to 10 tonal sounds per individual, when performed using a camcorder (Canon Legria FS200, 20009 possible. To investigate the behavioural context of sound pro- digital zoom, Tokyo, Japan) positioned c. 30 cm from the duction, two analyses were performed independently on the front of the experimental glass tanks. Audio recordings, as same dataset. The first was the chi-square test (at 5% level of described above, were conducted simultaneously, enabling significance) used to study the association of sound production sounds to be associated with specific behaviours. Male beha- with particular behavioural displays, in both thumps and tonal viours and the associated sound emissions were observed dur- sounds. Males courted females by displaying various beha- ing 17 recording sessions (2.8 per male) and analysed using vioural displays: behaviours expressed by the males were Windows Movie Maker (size: 1004 bytes; Microsoft, Red- quantified as the frequency of events, or the number of mond, WA, USA). Behavioural categories expressed by the observed behavioural acts occurring in a given time (n minÀ1). males were classified and scored according to our observa- Six behavioural categories were identified from the videos and tions and the existing behavioural literature for gobioids analysed: leading, chase, nest display, frontal display, upside- (Takemura, 1984; Amorim & Neves, 2007; Malavasi et al., down and pre-spawning. Sounds from vocal individuals were 2009, 2012). pooled, and the number of times each behavioural category was performed with or without sound emissions was counted. Adjusted residuals from the chi-square test were then used to Signal processing and sound analysis assess which behavioural categories were positively or nega- Analog sounds were digitalized at 4000 sampling rate (128 tively associated with the emission of sounds; that is, residuals accuracy, 16 bit resolution) and high-pass filtered (0.06 kHz) indicated whether the frequency of the cell was respectively to eliminate background noise. Additionally, +10 dB amplifica- overrepresented or underrepresented in the sample compared to tion was applied for better auditory and visual inspection of the expected frequency (Sebastianutto et al., 2008). The sec- the audio tracks. Digitalized sounds were analysed using Avi- ond analyses involved dividing the reproductive sequence into soft-SASLab Pro 5.2 Software (512-point fast Fourier trans- two phases (courtship and pre-spawning), and the number of form FFT, Hamming window; Avisoft Bioacoustics, Berlin, sounds recorded per phase was noted, separately for each Germany). Only sounds with a good signal to noise ratio male. As previously mentioned, courtship and pre-spawning (SNR) were analysed. Temporal features were measured from phases were recognized according to female absence or pres- the oscillogram, while frequency variables were obtained from ence in the nest, respectively. For the comparison between the the logarithmic power spectra (Hamming window, 512-points frequency of calls produced by each male during the courtship FFT, resolution 7 Hz). More rarely, thumps were produced in or pre-spawning phase, the t-test was applied for dependent sequences of several sounds, that is thump bursts, for which samples (5% level of significance) after logarithmic transfor- the sound parameters were also calculated. Visual inspection in mation of data. In addition, individual mean values of sound

Journal of Zoology 309 (2019) 269–279 ª 2019 The Zoological Society of London 271 Perccottus glenii sounds and vocal mechanism S. Horvatic et al. parameters were tested for correlation using the Pearson corre- using hand nets. All fish were measured, euthanized with an lation to investigate their mutual relationships and relationship overdose of MS-222 (tricaine methane sulphonate; Pharmaq, with physical characteristics (size and water temperature). Data Overhalla, Norway), and stored for 1 week in 7% formalde- were analysed with STATISTICA 13 (Statsoft Inc., Tulsa, OK, hyde fixative solution and then transferred to 70% ethanol. USA). Dissections of specimens were performed at the University of Liege (Belgium). Fish were dissected and examined with a Wild M10 binocular microscope (Leica Camera, Leica, Wet- Morphology zlar, Germany) equipped with a camera lucida. Dissections Eight males (six vocal and two randomly selected from stock concern mainly muscles in relation to the pectoral girdle fi aquaria, size range: 84–96 mm LS) were removed from tanks because of previous work implicating the pectoral n motion in sound production. The nomenclature used to designate parts of the musculature is based on previous studies (Winterbottom, 1974; Adriaens, Decleyre & Verraes, 1993).

Results

Acoustic properties of reproductive calls Sound production was documented in six P. glenii males during reproductive interactions (courtship and pre-spawning). Males were acoustically active in the presence of females, mostly when they approached the nest entrance. A total of 734 sounds were documented, and 171 were analysed (mean Æ SD = 15.3 Æ 6.8 thumps per male; 8.6 Æ 3.2 thump bursts per male; and 9.0 Æ 1.0 tonal sounds per male). Thump sounds and thump bursts (long trains of thumps) were produced by all males though tonal sounds were produced by three males only. During male–male aggressive interactions, all males were mute and no sounds were documented. Thumps are low-frequency sounds (~100 Hz) composed of irregularly repeated sound cycles of a relatively short duration (~95 ms; Fig. 1a). On the waveform, cycles varied in amplitude but the highest peak was the one in the first two-thirds of the sound. The first individual cycle had one peak, while the following cycles exhibited smaller extra-peaks, which again disappeared as the sound ceased (Fig. 1b). On the power spectrum, fundamental frequency was always the dominant frequency representing the peak with greatest energy, but sometimes, additional harmonics were observed. Call energy was decreased by c. À30 dB (relative intensity) by 600 Hz and did not exceed 1 kHz (Fig. 1c, Table 1). During reproductive interactions, males produced a mean of 9 thumps minÀ1, increasing vocal activity up to 19 thumps minÀ1 when the female approached the nest. Thumps were organized in long trains, that is sound bursts, in which thumps were spaced with a time interval of c. 1.2 s when the female hesitated on the nest entrance or was inspecting the cavity (Fig. 1a, Table 1). Bursts were composed from 2 to 11 thumps and in some cases lasted more than 10 s. Within the burst, thumps had the same Figure 1 a) Oscillogram and spectrogram of the thump burst, b) structure and temporal organization as when produced as indi- enlarged oscillogram of an individual thump and c) power spectrum vidual thumps; that is, they were short duration and low-fre- of the isolated thump produced during reproductive interactions of quency signals. The number of thumps within the burst was 2 Perccottus glenii. On the spectrogram, the marked area represents positively correlated with burst duration (Pearson r = 0.97, an energy band (reaching up to 500 Hz) with the arrows on the P < 0.05; Fig. 2). Positive and significant correlations were power spectrum indicating two frequency peaks (at 120 and 340 Hz). detected between water temperature and three thump burst Marked lines on the lower oscillogram represent sound duration, i.e. parameters: number of cycles (Pearson r = 0.44, N = 6, P < the onset and the end of one thump. On the oscillogram, arrows 0.05), cycle repetition rate (Pearson r = 0.30, N = 6, P < 0.05) indicate extra-peaks. RA – relative amplitude. and peak frequency (Pearson r = 0.40, N = 6, P < 0.05).

272 Journal of Zoology 309 (2019) 269–279 ª 2019 The Zoological Society of London S. Horvatic et al. Perccottus glenii sounds and vocal mechanism

Perccottus glenii males also produced short tonal sounds. Table 1 Sound parameters of recorded thumps and thump burst These were produced inside the nest or from its opening dur- during reproductive interactions in six Perccottus glenii males ing courtship in the presence of the ripe female. These sounds Sound parameters Mean SD Range were never produced when the female entered the nest during Thumps the pre-spawning phase. Tonal sounds were short sinusoidal À1 – ~ Thump rate (sounds min ) 9.3 3.2 2 18.5 vocalizations ( 90 ms) composed of about 10 sound cycles, Duration (ms) 95.5 7.3 64.0–148.0 whose repetition rate ranged from 91 to 125 Hz (Fig. 3a,b). Number of cycles 8.7 0.9 6.3–13.1 ~ Peak frequency was low ( 120 Hz) and appeared as a narrow Peak frequency (Hz) 97.9 4.4 74–121.1 peak on the spectrogram, strongly overlapping with the values Thump bursts of the cycle repetition rate (Fig. 3c, Table 2). In all cases, peak Thump duration (ms) 99.9 32.8 66.1–157.0 frequency was the fundamental frequency. Energy ranged up Thump number of cycles 7.6 1.1 5.5–9.2 to 300 Hz and did not exceed 500 Hz. Frequency modulation Thump peak frequency (Hz) 95.0 9.2 81.6–108.9 was slightly upward, with values averaging around 3 Hz. Burst duration (s) 4.4 3.8 1.3–13.2 Sound rate was lower when the female was further from the Number of thumps 4.4 3.1 2–11.5 nest, but increased as the female approached the nest entrance. Thump interval (s) 1.2 0.2 0.9–1.5 In these situations, tonal sounds were organized in series, averaging six repeated sounds per minute (Fig. 3a, Table 2). Sound variables, such as sound rate and frequency modulation, corre- associated with sound production. Generally, the most fre- lated positively with water temperature (Pearson r = 0.80 and quently observed behavioural categories (irrespective of sound r = 0.40, respectively; both N = 3, P < 0.05). type) were as follows: Nest display (33%, n = 99), Frontal display (22%, n = 68) and Leading (21%, n = 66). Tonal sounds Behavioural context of sound production were most commonly recorded during Frontal display (75%, n = 28), while 62% of documented categories were associated During reproductive interactions, the male courted the female with sound production (23 of 37 behaviours; Fig. 5). Thumps by performing a series of breeding displays separated in the were most frequently observed during Nest display (35%, six behavioural categories (Fig. 4), while swimming or while n = 94) and Leading (24%, n = 65). Only 23% of behavioural supported on the substrate by the pelvic fins. Body and head categories were associated with thumps (62 of 267 behaviours; movements observed during the sound emission in vocal fishes Fig. 5). were similar for both thump and tonal calls, without any obvi- Sound production differed in its association with the beha- ous differences in terms of motions. During sound production, vioural categories. The emission of thumps was positively fish exhibited pectoral fin abduction, rapid thrust of the throat associated with Frontal display and Pre-spawning region, and fin erection and body undulation. There was no (v2(6) = 804.2, P < 0.001). A positive association was also sound production during actual spawning (egg laying). found between tonal sounds and Frontal display (v2(6) = 62.3, During the 17 recording sessions (2.8 per male), 304 beha- P < 0.001), while fewer tonal sounds than statistically expected vioural acts were identified and separated into the six beha- were observed in other categories (v2(6) = 62.3, P < 0.001). In vioural categories. In total, 85 behaviours (42%) were addition, there was a significant difference between courtship

Figure 2 Relationship between burst duration and the number of thumps in Perccottus glenii. The curve was ﬁtted by: y = 1.1354x – 0.0861 (r2 = 0.97, P < 0.05).

Journal of Zoology 309 (2019) 269–279 ª 2019 The Zoological Society of London 273 Perccottus glenii sounds and vocal mechanism S. Horvatic et al.

test = 34.02, d.f. = 2, P < 0.001). For thumps, there was no sig- niﬁcant difference between the courtship or pre-spawning phases and the frequency of produced calls (courtship = 63.5 Æ 47.9 vs. pre-spawning = 54.6 Æ 73.0, dependent t-test = 1.29, d.f. = 5, P > 0.05).

Anatomical findings In previous studies on the sound-producing mechanisms in Gobiidae and Gobionellidae (Parmentier et al., 2013, 2017a), anatomical studies mainly concerned the pectoral girdles and associated muscles, that is levator pectoralis muscles, whose contractions were thought to be responsible for pectoral girdle motion during sound emission. In P. glenii, the pectoral girdle is articulated on the skull by the posttemporal bone. It consists of a basal plate with two processes that form a fork, prevent- ing forward displacement of the posttemporal bone (Fig. 6). The dorsal process articulates dorsally on the skull at the level of the epiotic. The ventral process is situated beneath the dorsal process and extends rostrally into a ligament attached to the intercalar bone. The supracleithrum connects the posttemporal to the cleithral bone. It is anteriorly covered by the posttemporal, but is caudally lateral to the dorsal process of the cleithrum. Baudelot’s ligament inserts on the medial face of the supracleithrum, bypasses caudally the anterior process of the cleithrum and inserts laterally on the basioccipital. Muscles described as sound-producing muscles were found in this part of the body. The levator pectoralis muscle is easily distinguishable and is separated into three bundles: a pars lateralis superficialis, a pars lateralis profundus and a pars medialis (Fig. 6). The musculus levator pectoralis pars lateralis superficialis originates from the dorsal caudal part of the pterotic bone of the neurocranium and inserts on the medial face of the posttemporal, at the level of the basal plate. Some fibres also insert on the ventral process of the posttemporal. The musculus levator pectoralis pars lateralis profundus originates on the caudal margin of the pterotic bone, rides ventrally along the ventral process of the posttemporal and inserts on the rostral margin of the cleithral bone. The musculus levator pectoralis pars medialis originates more medially, on the exoccipital bone, and is attached to the rostral side of the cleithral bone. On the clei- Figure 3 a) Oscillogram and spectrogram of the sinusoidal tonal thral bone, the insertion of the pars medialis is dorsal to pars sounds, b) enlarged oscillogram of an individual tonal sound and c) lateralis. power spectrum of the isolated tonal sound produced during reproductive interactions (courtship) of Perccottus glenii. On the Discussion spectrogram and power spectrum, the marked area represents the energy band (reaching up to 500 Hz) with the arrows on the The present study provides a qualitative and quantitative power spectrum indicating important frequency peaks (at 120 and assessment of the acoustic behaviour in the basal gobioid 470 Hz). Marked lines on the lower oscillogram represent sound P. glenii during reproduction and some insights into the puta- duration, i.e. the onset and the end of one tonal sound. RA – tive sound emission mechanism. Perccottus glenii males pro- relative amplitude. duced two different sound types, thumps and tonal sounds, during courtship and pre-spawning interactions. We did not record sounds during male intrasexual aggressive interactions, and pre-spawning in the frequency of recorded sounds only in what is interesting result considering high territoriality and the case of the tonal calls, where sounds were produced more breeding competition observed in P. glenii males and gobioids frequently during courtship than pre-spawning (mean Æ SD; in general (Takemura, 1984; Lugli et al., 1997; Amorim & courtship = 9 Æ 1 vs. pre-spawning = 1 Æ 0, dependent t- Neves, 2008). Thumps were the most frequent sound type and

274 Journal of Zoology 309 (2019) 269–279 ª 2019 The Zoological Society of London S. Horvatic et al. Perccottus glenii sounds and vocal mechanism

Figure 4 Behaviours associated with the sound production during reproductive interactions in Perccottus glenii. Descriptions: 1. Leading: a courtship behaviour where the male completely leaves the nest, approaches the female and then rapidly returns to the reproductive site; 2. Chase: male vigorously chases the female from the nest over a short (> 5 cm) or long (< 5 cm) distances while constantly biting her and producing sounds; 3. Nest display: male is completely inside the nest while the female swims around and away from the entrance, i.e. there is no visual contact; 4. Frontal display: male is orientated toward the female from the nest entrance (there is visual contact), and the female is positioned less than five centimetres from the nest opening, but not inside; 5. Upside-down: male turns upside down in the nest so that its genital papilla comes in contact with the nest ceiling; 6. Pre-spawning: while the female is completely inside the nest, the male undulates the body and waves the pectoral fins with erected dorsal fins. Resident males are black and intruder females grey. Nest and hydrophone size not to scale. were produced during both courtship and pre-spawning, while Lugli, 2000; Amorim & Neves, 2007; Malavasi et al., 2009), the female was outside the nest or within the cavity. Acousti- sounds of P. glenii were organized in long trains (sound cally, thumps were short duration (<200 ms) and low-fre- bursts) composed from 2 to 12 thump sounds, whose mean quency (<500 Hz) sounds, and exhibited an irregular acoustic parameters were similar to those of individually pro- waveform structure. Similar to sand gobies (Lindstrom€ & duced thumps. This study provides the first record of thump production outside the Gobionellidae, highlighting the com- Table 2 Sound parameters of recorded tonal sounds during plexity of acoustic signals within gobioids. reproductive interactions in three Perccottus glenii males We also recorded tonal-like sounds, which were only previously observed in a small proportion of European gobiids from Sound parameters Mean SD Range the Gobius lineage and Periophthalmodon septemradiatus Tonal sounds (Ladich & Kratochvil, 1989; Lugli et al., 1996, 1997; Malavasi À1 Sound rate (sounds min ) 6.4 1.2 4.7 to 7.7 et al., 2008; Sebastianutto et al., 2008; Polgar et al., 2011; Duration (ms) 89.7 16.8 70.4 to 119.5 Agorreta et al., 2013; Horvatić et al., 2015). In the case of Number of cycles 9.7 1.4 8.3 to 12.3 P. glenii, tonal sounds on the spectrogram were almost flat Cycle repetition rate (Hz) 111.1 10.4 91.7 to 123.7 non-modulated calls of short duration (<100 ms) with one Frequency modulation (Hz) 2.4 15.0 –25.0 to 20.4 important frequency peak (~120 Hz) corresponding to the Peak frequency (Hz) 117.1 4.5 112.3 to 124 cycle repetition rate, indicating its sinusoidal quality.

Journal of Zoology 309 (2019) 269–279 ª 2019 The Zoological Society of London 275 Perccottus glenii sounds and vocal mechanism S. Horvatic et al.

Figure 5 Percent of documented courtship behavioural acts (dark grey; total number of recorded acts; for thumps (n = 267) and tonal (n = 37) sounds) and the percentage of acts accompanied with the production of a) thump or b) tonal sounds (light grey) during courtship behaviour of Perccottus glenii males.

Perccottus glenii tonal calls show many common features with influences muscle contraction properties (Feher, Waybright & previously documented short tonal sounds described in certain Fine, 1998), and in our case, thumps tended to have a higher gobiids, that is Neogobius fluviatilis and Padogobius nigricans number of cycles, cycle repetition rate and peak frequency in (Lugli et al., 1996; Horvatić et al., 2015), highlighting the relation to higher temperatures within the burst. Tonal sounds spectro-temporal acoustic complexity (short, non-modulated vs. were more frequency modulated and produced with a higher long, frequency-modulated vocalizations) within gobioids, even emission rate with increasing temperature. Although prelimi- in the case of tonal sounds. In P. glenii, short tonal sounds nary, our findings are consistent with the hypothesis about were frequent during the courtship phase, that is, when females temperature-dependent sound-producing muscles in fishes (Ben- were inspecting the male territory. Though the small sample nett, 1985; Feher et al., 1998; Rome & Lindstedt, 1998). How- size hinders us from drawing deeper conclusions, the occur- ever, detailed study is needed in order to understand how rence of tonal sounds during the initial phase of the reproduc- temperature variation affects acoustic features in P. glenii. tive interactions (i.e. courtship) has been well documented in Male Padanian goby Padogobius bonelli produce tonal-like previous investigations of territorial nest-holding gobies calls separately from drumming sounds, or combine them to (Ladich & Kratochvil, 1989; Lugli et al., 1996; Horvatić et al., form complex sounds (Lugli et al., 1995, 1997). Though the 2015). Our results emphasize the complex situation about drumming and tonal components of complex sound never reproductive sequence-specific sound types in gobies. We pro- overlap, these components differ slightly in their acoustic pose that, in the case of P. glenii, tonal sounds could serve for parameters compared to the mean values of isolated sounds female attraction (i.e. when the female is outside the nest), (shorter and with a lower number of pulses). Therefore, it was while thump sounds are utilized in pre-spawning interactions, suggested that a single sonic mechanism could be responsible once the female has entered the nest hollow. for sound production in P. bonelli (Lugli et al., 1995). Varia- In this study, a relationship was found between water tem- tion rates of the same muscles could generate either discrete perature and acoustic parameters, in both sound types. Sound pulses (drums) or rapidly repeated pulses yielding tone-like characteristics are expected to change with temperature since it sounds. Our study demonstrates that P. glenii males were also

276 Journal of Zoology 309 (2019) 269–279 ª 2019 The Zoological Society of London S. Horvatic et al. Perccottus glenii sounds and vocal mechanism

Figure 6 Dorsal view of the putative sound-producing mechanism in Perccottus glenii. On the right side, the posttemporal, levator pectoralis lateralis superficialis and levator pectoralis medialis muscles were removed to see the levator pectoralis lateralis profundus muscle. [Colour figure can be viewed at zslpublications.onlinelibrary.wiley.com] able to produce two acoustically different sound types. How- to this, the oscillogram shape of tonal sounds is more constant, ever, the question remains as to how P. glenii thumps and providing less additional spectral components (Fig. 3c). In this tonal sounds are produced. No clear differences were observed case, tonal sounds could correspond to fully completed thumps. in the cranio-pectoral movements associated with thump or In other words, thumps and tonal sound could be produced by tonal sound production. Further, we did not observe nodding the same sound-producing mechanism involving the levator behaviour or forward girdle displacement, as previously sug- pectoralis contractions, though these contractions provide more gested for some cottids and gobies (Colleye et al., 2013; Zeyl regular cycles in tonal sounds. et al., 2016). Therefore, it is possible that the basal odontobu- According to Malavasi et al. (2008), interspecific compar- tid P. glenii possesses a single mechanism that is responsible ison of acoustic signals across eleven Mediterranean gobies for the production of both sound types. Detailed high-speed suggested that the pulsatile sound could represent an ancestral recordings and multidisciplinary research (muscle histology and state within the gobies. In respect to this hypothesis, the pre- electromyography) are needed to fully explain the sound emis- sent results broaden the evolutionary scenario regarding acous- sion process in P. glenii. tic communication among the gobioids. We suggest that Fine et al. (2001) demonstrated that muscle contraction rate pulsatile (thump-like) and short tonal-like sounds could consti- dictates the fundamental frequency of calls of the male oyster tute the fundamental acoustic unit within the gobioids, since toadfish (Opsanus tau), with values overlapping peak sound they are clearly documented in basal odontobutids (present amplitude. In our study, the comparison of oscillograms and study and Takemura, 1984). Although speculative, it is possi- power spectra between thump and tonal sounds provides inter- ble that these ‘stem’ sounds could have evolved in terms of esting information. Dominant frequencies in both thump and temporal complexity (i.e. temporal spacing of pulse repetition) tonal sounds mostly overlapped but thumps showed additional and frequency variation (modulation), presenting a diversity of spectral components at frequencies above 200 Hz (Figs 1c and long, frequency-modulated tonal and pulsatile/drumming sig- 3c). These additional components in thumps reflect the less nals reported currently for gobiids and gobionellids. By pro- regular sinusoidal shape of the oscillogram (Fig. 1b). Contrary ducing these fundamental units and additionally combining

Journal of Zoology 309 (2019) 269–279 ª 2019 The Zoological Society of London 277 Perccottus glenii sounds and vocal mechanism S. Horvatic et al. them, some gobiids (i.e. P. bonelli and Gobius cruentatus) Amorim, M.C.P. & Neves, A.S.M. (2008). Male painted goby have the ability to emit complex calls possessing both tonal (Pomatoschistus pictus) vocalise to defend territories. and drumming components within the same burst. Behaviour 145, 1065–1083. Bennett, A.F. (1985). Temperature and muscle. J. Exp. Biol. 115 – Conclusions , 333 344. Blom, E.L., Muck,€ I., Heubel, K. & Svensson, O. (2016). This study supports the hypothesis that sound production and Acoustic and visual courtship traits in two sympatric marine acoustic communication evolved early within the gobioid lin- Gobiidae species — Pomatoschistus microps and eage. Thumps and short tonal-like sounds reported for the Pomatoschistus minutus. Environ. Biol. Fishes 99, 999–1007. odontobutid P. glenii could be considered a synapomorphic Colleye, O., Ovidio, M., Salmon, A. & Parmentier, E. (2013). trait for gobioid fishes. However, additional experimental study Contribution to the study of acoustic communication in two is required to determine whether sounds are produced in other Belgian river bullheads (Cottus rhenanus and C. perifretum) basal gobioid families (mainly Eleotridae and Butidae) and to with further insight into the sound-producing mechanism. be able to compare them with other taxa. In addition, we pro- Front. Zool. 10, 71. pose that muscles originating on the skull and attaching to the Feher, J.J., Waybright, T.D. & Fine, M.L. (1998). Comparison of pectoral girdle have an important role in the sound emission of sarcoplasmic reticulum capabilities in toadfish (Opsanus tau)sonic P. glenii and therefore present a putative mechanism involved muscle and rat fast twitch muscle. J. Muscle Res. 19,661–674. in their acoustic communication. In summary, this study Fine, M.L., Malloy, K.L., King, C., Mitchell, S.L. & Cameron, expands the knowledge about gobioid vocal behaviour and T.M. (2001). Movement and sound generation by the toadfish underlines the importance of acoustic communication within J. Comp. Physiol. A Sens. Neural. Behav. this group of fish. swimbladder. Physiol. 187, 371–379. Freyhof, J. (2011). Diversity and distribution of freshwater Acknowledgements gobies from the Mediterranean, the Black and Caspian seas. The authors are grateful to A. Speli c, D. Hamer and K. Horvat In The biology of gobies: 279–288. Patzner, R.A., Van for their kind and useful assistance during the laboratory audio- Tassell, J.L., Kovacić, M. & Kapoor, B.G. (Eds). Enfield: video recordings. We thank S. Sviben for initial paper revision Science Publishers. and L. Zanella for proofreading the paper. In addition, we thank Horvatić, S., Cavraro, F., Zanella, D. and Malavasi, S. M. Vuckovic for figure preparation regarding behavioural acts. 2015.Sound production in the Ponto-Caspian goby Neogobius Lastly, authors are thankful to Michael L. Fine and an anony- fluviatilis and acoustic affinities within the Gobius lineage: mous reviewer whose comments significantly improved the ini- implications for phylogeny. Biol. J. Linn. Soc. 117, 564–573. tial version of the paper. This research was funded by Ministry Kottelat, M. & Freyhof, J. (2007). Handbook of European of Education, Science and Technological Development, Serbia freshwater fishes. Kottelat, Comol, Switzerland andBerlin, (projects TR37009 and IO173045) and Ministry of Environment Germany: Freyhof. and Energy, Croatia (project 7055462). Ladich, F. & Fine, M.L. (2006). Sound-generating mechanisms in fishes: a unique diversity in vertebrates. In Communication fi – Conflict of interest in shes:343. Ladich, F., Collin, S.P., Moller, P. & Kapoor, B.G. (Eds). Enfield: Science Publishers. None. Ladich, F. & Kratochvil, H. (1989). Sound production by the marmoreal goby Proterorhinus marmoratus (Pallas) (Gobiidae, 93 – References Teleostei). Zool. Jahrb. , 501 504. Lindstrom,€ K. & Lugli, M. (2000). A quantitative analysis of the Adriaens, D., Decleyre, D. & Verraes, W. (1993). Morphology courtship acoustic behaviour and sound patterning in male of the pectoral girdle in Pomatoschistus lozanoi De Buen, sand goby, Pomatoschistus minutus. Environ. Biol. Fish. 58, 1923 (Gobiidae), in relation to pectoral fin adduction. Belg. J. 411–424. Zool. 123, 137–157. Lugli, M., Pavan, G., Torricelli, P. & Bobbio, L. (1995). Agorreta, A., San Mauro, D., Schliewen, U., Van Tassell, J.L., Spawning vocalizations in male freshwater gobiids (Pisces, Kovacic, M., Zardoya, R. & Ruber,€ L. (2013). Molecular Gobiidae). Environ. Biol. Fish. 43, 219–231. phylogenetics of Gobioidei and phylogenetic placement of Lugli, M., Torricelli, P., Pavan, G. & Miller, P.J. (1996). European gobies. Mol. Phylogenet. Evol. 69, 619–633. Breeding sounds of male Padogobius nigricans with Akamatsu, T., Okumura, T., Novarini, N. & Yan, H.Y. (2002). suggestions for further evolutionary study of vocal behaviour Empirical refinements applicable to the recording of fish in gobioid fishes. J. Fish Biol. 49, 648–657. sounds in small tanks. J. Acoust. Soc. Am. 112, 3073–3082. Lugli, M., Torricelli, P., Pavan, G. & Mainardi, D. (1997). Amorim, M.C.P. & Neves, A.S.M. (2007). Acoustic signalling Sound production during courtship and spawning among during courtship in the painted goby, Pomatoschistus pictus. freshwater gobiids (Pisces, Gobiidae). Mar. Freshw. Behav. J. Mar. Biol. Assoc. U.K. 87, 1017–1023. Physiol. 29, 109–126.

278 Journal of Zoology 309 (2019) 269–279 ª 2019 The Zoological Society of London S. Horvatic et al. Perccottus glenii sounds and vocal mechanism

Malavasi, S., Collatuzzo, S. & Torricelli, P. (2008). Interspecific Parmentier, E., Diogo, R. & Fine, M.L. (2017b). Multiple variation of acoustic signals in Mediterranean gobies exaptations leading to fish sound production. Fish Fish. 18, (Perciformes, Gobiidae): comparative analysis and 958–966. evolutionary outlook. Biol. J. Linn. Soc. 93, 763–778. Polgar, G., Malavasi, S., Cipolato, G., Georgalas, V., Clack, J.A. Malavasi, S., Valerio, C. & Torricelli, P. (2009). Courtship & Torricelli, P. (2011). Acoustic communication at the water's sounds and associated behaviours in the Canestrini's edge: evolutionary insights from a mudskipper. PLoS ONE 6, goby Pomatoschistus canestrinii. J. Fish Biol. 75, 1883–1887. e21434. Malavasi, S., Gkenas, C., Leonardos, I., Torricelli, P. & Rome, L.C. & Lindstedt, S.L. (1998). The quest for speed: McLennan, D.A. (2012). The phylogeny of a reduced ‘sand muscles built for high-frequency contractions. News Physiol. goby’ group based on behavioural and life history characters. Sci. 13, 261–268. Zool. J. Linn. Soc. 165, 916–924. Sebastianutto, L., Picciulin, M., Costantini, M., Rocca, M. & Miller, P.J. (1986). Gobiidae. In Fishes of the North-Eastern Ferrero, E. (2008). Four types of sounds from one winner: Atlantic and the Mediterranean: 1019–1085. Whitehead, vocalizations during territorial behaviour in the red-mouthed P.J.P., Bauchot, M.L., Hureau, J.-C., Nielsen, J. & Tortonese, goby Gobius cruentatus (Pisces, Gobiidae). Acta Ethol. 11, E. (Eds). Paris: UNESCO. 115–121. Miller, P.J. (2003). Gobiidae I. In The freshwater fishes of Takemura, A. (1984). Acoustical behaviour of the freshwater Europe: 157–398. Miller, P.J. (Ed). Wiesbaden: Aula-Verlag. goby Odontobutis obscura. Bull. Jpn. Soc. Sci. Fish. 50, 561– Myrberg, A.A. & Lugli, M. (2006). Reproductive behavior and 564. acoustical interactions. In Communication in fishes: 149–176. Thacker, C.E. (2009). Phylogeny of Gobioidei and placement Ladich, F., Collin, S.P., Moller, P. & Kapoor, B.G. (Eds). within Acantomorpha with a new classification and Enfield: Science Publishers. investigation of diversification and character evolution. Copeia Parmentier, E. & Fine, M.L. (2016). Fish sound production: 2009,93–104. insight. In Vertebrate sound production and acoustic Winterbottom, R. (1974). A descriptive synonymy of the striated communication:19–49. Suthers, R., Tecumseh, F., Popper, muscles of the Teleostei. Proc. Acad. Nat. Sci. Philad. 125, A.N. & Fay, R.R. (Eds). New York: Springer. 225–317. Parmentier, E., Kever, L., Boyle, K., Corbisier, Y., Sawelew, L. Zeyl, J.N., Malavasi, S., Holt, D.E., Noel, P., Lugli, M. & & Malavasi, S. (2013). Sound production mechanism in Johnston, C.E. (2016). Convergent aspects of acoustic Gobius paganellus (Gobiidae). J. Exp. Biol. 216, 3189–3199. communication in darters, sculpins and gobies. In In fish Parmentier, E., Petrinisec, M., Fonseca, P.J. & Amorim, M.C.P. hearing and bioacoustics: an anthology in honor of Arthur N. (2017a). Sound-production mechanism in Pomatoschistus Popper and Richard R. Fay:93–120. Sisneros, A.J. (Ed.). pictus. J. Exp. Biol. 220, 4374–4376. Switzerland: Springer.

Journal of Zoology 309 (2019) 269–279 ª 2019 The Zoological Society of London 279

1 Correlation between acoustic divergence and phylogenetic distance in vocal

2 European gobiids (Gobiidae; Gobius lineage)

3 Relationship between sound variability and genetic structure in gobies

5 Sven Horvatić1, Stefano Malavasi2, Jasna Vukić3, Radek Šanda4, Zoran Marčić1, Marko Ćaleta5,

6 Massimo Lorenzoni6, Perica Mustafić1, Ivana Buj1, Lucija Raguž1, Lucija Ivić1, Francesco Cavraro2,

7 Davor Zanella1*

9 1 Department of Zoology, Faculty of Science, University of Zagreb, Zagreb, Croatia

10 2 Department Environmental Sciences, Informatics and Statistics, Ca` Foscari, University of

11 Venice, Venezia Mestre, Italy

12 3 Department of Ecology, Charles University, Prague, Czech Republic

13 4 National Museum, Department of Zoology, Prague, Czech Republic

14 5 Faculty of Teacher Education, University of Zagreb, Zagreb, Croatia

15 6 Department of Chemistry, Biology and Biotechnologies, University of Perugia, Perugia, Italy

17 * Corresponding author

18 E-mail: [email protected] (DZ)

Predano na recenziju u Plos One

19 Abstract

20 In fish, identity can be encoded by sounds, which have been thoroughly investigated in

21 European gobiids (Gobiidae, Gobius lineage). Recent evolutionary studies suggested that

22 deterministic and/or stochastic forces could generate acoustic differences among related animal

23 species, but this was never investigated in any Teleost group up to date. In the present

24 comparative study, we analysed the sounds from nine vocal gobiids and quantitatively assessed

25 their acoustic variability. Overall, we analysed the means of 616 sounds from 67 males. Our

26 interspecific acoustic study, incorporating for the first time the acoustic signals from the majority

27 of vocal gobiids, suggested that their sounds are truly species-specific (92% of sounds were

28 correctly classified into exact species) and that each taxon possesses a unique set of spectro-

29 temporal variables. In addition, we reconstructed phylogenetic relationships from a concatenated

30 genetic dataset consisting of multiple genetic markers to track the evolution of acoustic

31 communication in vocal gobiids. The results of this study indicated that the genus Padogobius is

32 polyphyletic, since P. nigricans was nested within the Ponto-Caspian clade, while the congeneric

33 P. bonelli turned out to be a sister taxon to the remaining vocal species. Lastly, by extracting the

34 acoustic and genetic distance matrices, sound variability and genetic distance were correlated for

35 the first time to assess whether sound evolution follows the similar phylogenetic pattern.

36 Significance of sounds in goby evolution, invoked by many previous studies, was empirically

37 corroborated by the positive correlation between the sound variability and genetic distance,

38 emphasizing that some acoustic features could carry the phylogenetic signal important for

39 accurate species-discrimination in vocal gobiids. Positive correlation supported the hypothesis

40 that sound variability in nine European gobiids is under the influence of stochastic forces. Our

41 study was the first attempt to evaluate the mutual relationship between acoustic variation and

42 genetic divergence in any Teleost fish.

Predano na recenziju u Plos One

44 Introduction

45 Many animal species use sounds, together with other communication signals, to express their

46 behaviour, and by actively changing their acoustic properties, they can control the information

47 content of these signals [1,2]. Interspecifically, sounds can encode the identity of the signalling

48 individual [3,4,5] and for related species in sympatry, this is an important discrimination trait

49 during reproductive interactions [6,7,8]. Given the significant role of sound production in the

50 species recognition process, it is believed that divergence in acoustic signals could drive

51 speciation [9-12].

52 One of the central questions in bioacoustics, when it comes to divergence caused by variations

53 in acoustic signals, is to determine which evolutionary forces have generated prominent

54 interspecific differences among animal taxa. Even though a signal evolution is rarely explained by

55 a single evolutionary force, most of the studies invoke two common processes or forces which are

56 generally thought to be responsible for acoustic divergence: deterministic and stochastic.

57 Deterministic (or ‘adaptive’) processes, such as habitat adaptation [13,14], divergence in

58 morphology [15,16] or selection for species recognition (“reproductive character displacement”

59 or sometimes even the sexual selection) [17,18,19] act to amplify the signal variations already

60 present among species. These processes generate straightforward predictions about the

61 direction of evolution [20]. In these circumstances, an absence of association between genotype

62 and acoustics highlights the importance of deterministic factors and other selective pressures in

63 shaping acoustic traits. On the other hand, stochastic (or ‘neutral’) processes, such as sexual or

64 social selection [21,22] or more commonly, genetic drift and mutation [13,23,24,25] could be the

65 driving forces for signal divergence. These mechanisms make signal divergence a highly

66 stochastic and unpredictable process [20], where a positive correlation is usually observed

67 between call divergence and genotype. Likewise, some studies indicate that the two processes

68 sometimes interact and act mutually, causing the signal divergence [for references, see 20].

Predano na recenziju u Plos One

69 Accordingly, studies on animal sounds, supporting both the association between acoustic

70 variation and genetic divergence [20,25-30] and the lack of association between the two patterns

71 [13,31,32].

72 The vocal repertoire among Teleosts has been thoroughly investigated in the gobioids

73 (Gobiiformes; Gobioidei). Soniferous species produce different types of acoustic signals,

74 presenting a rich repertoire composed from pulsatile to tonal sounds. The acoustic structure

75 shows great variability at both the inter- and intraspecific levels, with four different sound types

76 [thump, pulsatile (drum), tonal, and complex] recorded to date, emitted mainly by males as part

77 of the breeding and aggressive behavioural repertoire. Specifically, since sounds are produced

78 during reproduction for inter- and intrasexual interactions like in, for example, painted goby

79 (Pomatoschistus pictus (Malm, 1865)) and monkey goby (Neogobius fluviatilis (Pallas, 1814)) [33-

80 35], they are important for evolutionary studies examining the processes leading to species

81 radiation. In addition, due to the morphological similarities and the lack of sonic specializations

82 in currently investigated species [36,37], it can be expected that gobies utilize similar acoustic

83 components for vocal communication which could reveal a certain phylogenetic pattern.

84 Taxonomically, the suborder Gobioidei is one of the largest vertebrate groups, including

85 several families [38-40]. The European gobies belong to two of these families, Gobiidae and

86 Gobionellidae [41]. Within the European gobies, three lineages have been recognized, with the

87 Gobius- and Aphia-lineage as part of the family Gobiidae (gobiine-like clade in [42]), while the

88 Pomatoschistus-lineage was nested within the family Gobionellidae (gobionelline-like clade in

89 [42]). Traditionally, molecular studies have strongly emphasized that the European gobies from

90 the Gobius-lineage form a monophyletic group [40,42,43], whereas the genus Gobius (including

91 Zosterisessor; [40,42,44,45] and endemic goby species from the Ponto-Caspian region (e.g. genera

92 Babka, Benthophilus, Mesogobius, Neogobius Proterorhinus, Ponticola) [40,42,46] are by far the

93 most species-rich groups. Up to date, sound production was most commonly documented within

94 Gobius lineage, for which 11 species proved to be soniferous during the experimental trials (8).

Predano na recenziju u Plos One

95 From the total number of vocal gobiids (Gobiidae), this makes around 73% of all soniferous taxa

96 and highlights the fact that the sound production is a common phenotypic trait for Gobius lineage.

97 On the contrary, only eight gobionellids (Gobionellidae) were vocally active during the

98 bioacoustic experiments, and the majority of taxa, i.e. six species, belonged to Pomatoschistus

99 lineage (8). Lastly, only two species of odontobutids (Odontobutidae) were tested for sound

100 production (47,48), emphasizing that the acoustic communication is rarely investigated outside

101 Gobiidae and Gobionellidae. In Gobius lineage, interspecific sound diversification is thought to be

102 highly important, since closely related species show less overlap of their call features and have

103 even developed different kinds of calls. It was proposed that those differences follow the

104 phylogenetic pattern to certain degree [34,48,49,50], however, these studies were empirically

105 limited, since no one has compared acoustic signals with genetic data to corroborate possible

106 concordance. Furthermore, these studies have supported findings from other fish groups [51-56],

107 showing that some acoustic features could be a reliable species or individual identifier. Likewise,

108 recently documented acoustic communication in Amur sleeper Perccottus glenii (Odontobutidae)

109 suggests a deeper sound production ancestry within the gobioids [48]. Since gobies are widely

110 distributed in European waters [57-59] and many species live in sympatry with at least one other

111 species, communication signals (including sounds) likely play a significant role in mating

112 recognition and prevention of hybridization. Therefore, the observed vocal diversity and sound

113 utilization during reproduction indicate that acoustic communication could have a prominent

114 role in the evolution and speciation of the European gobies. To date, no robust, comparative or

115 in-depth studies incorporating a correlation between acoustic signals and genetic markers have

116 been conducted on the acoustic communication between closely related species in Gobiidae.

117 Therefore, the object of this study was to give a quantitative evaluation of the relationship

118 between interspecific acoustic variation and genetic divergence in vocal Gobius lineage gobies

119 and, according to the observed association, discuss about the potential evolutionary forces

120 promoting the sound divergence.

Predano na recenziju u Plos One

121 By combining quantitative bioacoustics and multiple genetic markers, this study examines

122 previously documented but never comprehensively analysed acoustic diversity of the sounds

123 produced by nine gobiid species (Gobiidae, Gobius lineage) and explores the degree to which the

124 affinities in acoustic signals between closely related species could be related to their phylogenetic

125 relationships. Specifically, this study aimed to: i. explore the interspecific acoustic variation

126 among nine Gobius lineage gobiids and assess the acoustic variables responsible for the species

127 differentiation; ii. investigate the phylogenetic relationship between vocal Gobius lineage gobiids;

128 iii. examine the correlation between interspecific acoustic divergence and genetic distance based

129 on multilocus data (two mtDNA and two nDNA markers) to explore the phylogenetic significance

130 of acoustic signals in species diversity; iv. obtain insight into the evolutionary forces driving

131 acoustic divergence in these taxa, and v. construct phylogenetic hypotheses on the evolution of

132 acoustic communication in soniferous gobioids.

133

134 Material and methods

135 Study species

136 This study analysed acoustic signals and compared the vocal communication of nine vocal

137 gobiids (Gobiidae, Gobius lineage) belonging to five genera (Gobius, Padogobius, Zosterisessor,

138 Neogobius & Ponticola). The sounds were previously recorded and described by the authors of

139 the present study (see ‘Sound recording and bioacoustic analyses’) but were never assembled

140 into single comprehensive phylogenetic framework. Therefore, our species composition was

141 based on the availability of audio tracks for interspecific acoustic analysis. Amur sleeper

142 Perccottus glenii is a vocal Eurasian gobioid belonging to the family Odontobutidae or the

143 “sleepers”. Due to the sister phylogenetic position of sleepers to the rest of the gobies including

144 Gobiidae [41,42], P. glenii served as an outgroup in the analyses. The nine investigated goby

Predano na recenziju u Plos One

145 species belong to the family Gobiidae [41-43], more specifically, to the Gobius lineage sensu [42].

146 Traditionally, these nine gobiids have been split into two groups, Atlantic-Mediterranean (AM)

147 and Ponto-Caspian (PC) [59]. In the study, the AM gobies include the genera Gobius, Zosterisessor

148 and Padogobius. Among them, black goby Gobius niger Linnaeus, 1758, giant goby G. cobitis Pallas,

149 1814, rock goby G. paganellus Linnaeus, 1758 and grass goby Zosterisessor ophiocephalus (Pallas,

150 1814) are marine/brackish inhabitants. Gobius niger usually occupies similar muddy habitat as Z.

151 ophiocephalus, while the two Gobius species (G. paganellus and G. cobitis) appear sympatric in

152 rocky habitats (pers. obs.). Two Padogobius species (Padanian goby Padogobius bonelli

153 (Bonaparte, 1846) and Arno goby P. nigricans (Canestrini, 1867) are pure freshwater Italian

154 endemics, while P. bonelli also occurs in Croatian rivers. Gobius paganellus, G. cobitis and two

155 Padogobius species usually occupy stony/pebble substrates, while Gobius niger and Zosterisessor

156 ophiocephalus can be found on sandy/muddy bottoms [59,60]. The PC species are mostly brackish

157 to freshwater residents; bighead goby Ponticola kessleri (Günther, 1861) occupies stony or gravel

158 habitat, similarly to monkey goby Neogobius fluviatilis (Pallas, 1814) which is common on gravel

159 or sandy substrates. Round goby Neogobius melanostomus (Pallas, 1814) is common on a wide

160 range of substrates [57,58,61]. In several Croatian watercourses, PC gobies live sympatrically and

161 occupy similar bottom type (pers. obs.). The data regarding life-history traits (number of

162 vertebrae and swim bladder presence) were adopted from the available literature (57,58,62) or

163 from our field observations (habitat).

164

165 Genomic sampling and phylogenetic analyses

166 DNA was extracted from fin clips preserved in 96% ethanol using a Geneaid® DNA

167 Isolation Kit. Samples were amplified for mitochondrial genes cytochrome b (cytb) and

168 cytochrome c oxidase subunit I (cox1), and for nuclear genes Recombination activating gene 1

169 (rag1) and Rhodopsin (rho). Cytb and rag1 were amplified according to the protocol described in

Predano na recenziju u Plos One

170 [63] using either primers AJG and H5 [64] or GluF and ThrR [65] for cytb and RAG1F1 and

171 RAG1R1 [66] for rag1. Cox1 was amplified with primers FishF1 and FishR1 [67] according to the

172 PCR protocol of [68]. Rho was amplified with primers RhodF and RhodR [69]. The PCR protocol

173 was as follows: initial denaturation at 94°C for 5 min, followed by 35 cycles of denaturation,

174 annealing and elongation (94°C for 45 s, 53°C for 1 min and 72°C for 2 min), and the final

175 elongation at 72°C for 10 min. PCR products were purified with ExoSAP-IT and sequencing was

176 performed by Macrogen Europe (Netherlands) using amplification primers. The remaining

177 sequences were downloaded from GenBank [42,46,70] (Table 1). Sequences [S2] were visually

178 checked in Chromas v2.6.4 and aligned in Bioedit v7.2.6.1 [71]. New sequences were deposited in

179 GenBank (Table 1). The phylogenetic reconstruction analyses were conducted on a concatenated

180 dataset of all genes. Concatenation was recently evaluated as an appropriate method [72,73].

181 Prior to analysing the sequence data, the best fitting model of nucleotide substitution for each

182 marker and subset of positions inside the codons was determined by PartitionFinder 2 [74],

183 according to Bayesian information criterion (BIC) and under the all models option. The selected

184 partitioning scheme is listed in the Table S1. Bayesian Inference (BI) and Maximum Likelihood

185 (ML) approaches were used to estimate the phylogenetic relationships between the species. BI

186 was conducted in MrBayes v3.2.2 [75] with four independent MCMC runs for 2 million

187 generations, applying the partitioning scheme assessed by PartitionFinder 2. Trees were sampled

188 every 1000 generations. The convergence of runs was analysed and visualised in TRACER v1.7.0.

189 The first 25% of sampled trees were discarded as burn-in. The remaining trees were used to

190 construct a 50% majority-rule consensus tree. Randomized Axelerated Maximum Likelihood

191 [RAxML 8.2.12, 76] was used to assess ML, using Science Gateway portal CIPRES [77]. Partitioning

192 scheme assessed by PartitionFinder 2 was applied. Support of nodes was estimated by applying

193 1000 nonparametric bootstrap replicates. Genetic distances (uncorrected p-distances) were

194 assessed with MEGA 6 [78]. Nucleotide composition homogeneity within genes was tested with

195 PAUP* 4.0b10 [79].

196

Predano na recenziju u Plos One

197 Table 1. Location, geographic position and GenBank accession numbers of gobioids used in the

198 phylogenetic analyses.

Geographic position GenBank accession no. Taxon Location Latitude Longitude cytb cox1 rag1 rho Danube River, Veliko Gradište, Perccottus glenii SRB 44.763586 21.431633 this study AY722171 KF415837 KX224234 Gobius cobitis Adriatic Sea, Seline, CRO 45.265168 15.479553 this study KR914767 this study this study Gobius paganellus Adriatic Sea, Karinsko ždrilo, CRO 45.170532 15.598772 this study KR914777 this study this study Gobius niger / / / KF415583 KR914775 FJ526891 this study Zosterissesor ophiocephalus / / / EU444670 this study FJ526851 this study Padogobius nigricans Topino River, Umbria, IT 43.098344 12.765068 this study KJ554001 this study this study Padogobius bonelii Zrmanja River, Muškovci, CRO 44.208541 15.739067 this study KJ554527 this study this study Kupa-Kupa Channel, Donja Neogobius fluviatilis Kupčina, CRO 45.531749 15.789373 this study FJ526807 EU444718 this study Neogobius melanostomus Sava River, Mačkovac, CRO 45.748060 16.229190 this study FJ526801 FJ526857 JF261593 Ponticola kessleri / / / FJ526770 FJ526825 FJ526879 this study

199 Abbreviations: SRB, Serbia, CRO, Croatia, IT, Italy.

200 Sound recordings and bioacoustic analyses

201 For all investigated taxa, audio recordings were obtained from laboratory studies

202 [34,49,50,48,80,81,82], and the recording protocols and acoustic terminology were adopted as

203 closely as possible to allow for interspecific comparison. Our acoustic dataset consists of 67 vocal

204 Gobius lineage individuals (for nine species) for which at least ten calls were recorded per

205 individual and the individual means for each variable were calculated (mean ± SD = 87.0 ± 33.7

206 sounds analysed per species). Briefly, sounds were recorded from males under laboratory

207 conditions during the mating season (all gobies spawn from early spring to late summer), using

208 different audio equipment consisting of a hydrophone (Gulton Industries, HTI 94 SSQ or H2A-

209 XLR) with preamplifiers (B&K 2626 or IRIG PRE) connected to a portable audio recorder (ZOOM

210 H4n, Sony D7 or Tascam Linear PCM). Sounds were monitored and recorded during the “intruder

211 test”, where one individual exhibiting highly territorial behaviour after one week of

212 acclimatization to laboratory conditions was marked as the resident fish occupying the shelter,

213 in order to elicit inter- (male-female) or intrasexual (male-male) interactions. After recording,

Predano na recenziju u Plos One

214 sounds were digitized (.wav format) and analysed using AVISOFT SASLab Pro Software (version

215 5.2.14., Berlin, Germany) which allowed for calculation of the acoustic variables important for

216 further interspecific acoustic comparison. In addition, the spectrogram, oscillogram and power

217 spectrum were prepared using AVISOFT (Hamming window, 512-points FFT, resolution 7 Hz).

218 Only sounds with a good signal-to-noise ratio were used in the analysis. Most investigated species

219 produce only one type of sound, while some gobies (P. glenii and G. paganellus) produce two

220 sounds or even possess an elaborate vocal repertoire composed of three different calls (P. bonelli).

221 However, the criteria used in this study for all species implies only one representative sound type

222 per species for further comparative analysis. The representative sound type for each species was

223 selected based on the overall number of calls observed in the audio recordings, i.e. the sound type

224 that was most frequently registered and recorded during the behavioural investigations. To

225 investigate interspecific affinities, audio recordings obtained from previous studies were re-

226 analysed [34,49,50,48,80,81,82] and the six acoustic properties describing the temporal and

227 spectral structure of gobiid sounds were calculated. Temporal parameters were sound rate (SR,

228 sounds/min), number of pulses (NP), duration (DUR, milliseconds), and pulse repetition rate

229 (PRR, dividing number of pulses with duration, in hertz). Peak frequency (PF, highest peak in

230 power spectrum; hertz) and frequency modulation (FM) calculated as the difference between

231 final PRR and initial PRR and expressed in hertz, were spectral variables in our analyses. The main

232 purpose of the acoustic analysis was to construct a robust acoustic dataset ready for pairwise

233 comparison with genetic divergence.

234

235 Comparison between acoustic and genetic data

236 To assess whether the acoustic interspecific differences in gobiids were related to

237 phylogenetic relationships, we investigated the association between call divergence and genetic

238 distance using the Mantel test as a prior choice [83,84]. For the correlation, we used acoustic

Predano na recenziju u Plos One

239 distance matrix constructed from Cluster analysis (Joining tree analysis) in STATISTICA® (version

240 13.6.0., TIBCO, USA), in which clustering was performed with nine Gobiidae species as a grouping

241 variable and six acoustic features as the analysis variables (i.e. dimensions). For the

242 amalgamation (aggregation) rule, we used unweighted pair-group average (UPGMA) linkage,

243 while the distance matrix was computed from the means of all sound variables for each species

244 and built using the City-block (Manhattan) distance metric procedure. Genetic distance matrix

245 was assessed using the uncorrected p-distance method in MEGA (version 10.0.5., USA), based on

246 the concatenated data set for all used markers, and separately for mtDNA (cytb and cox1) and

247 nDNA markers (rag1 and rho). We used the bootstrap variance estimation method with 1,000

248 replications and p-distance as a substitution model for constructing the pairwise distance

249 between the taxa. The Mantel test was conducted in PASSaGE v.2 [85] on 10x10 distance matrices

250 with 10,000 permutations. Likewise, two additional Mantel tests were performed between the

251 obtained acoustic distance matrix and genetic distances based on 1) mtDNA (cytb and cox1) and

252 2) nDNA (rag1 and rho) markers. These correlations were performed to investigate the

253 relationship between call divergence and genetic distance using DNA markers with different rates

254 of mutation or evolution and therefore, they could reveal different aspects of the speciation

255 history of the examined taxa.

256

257 Statistical analyses

258 Descriptive statistics were calculated for each temporal (SR, NP, DUR, PRR) and spectral

259 (PF and FM) property of the acoustic signal produced by each species. For the preliminary

260 explorations, we considered all sound variables except in the case of P. glenii, where it was not

261 possible to calculate the FM of the thump sounds (i.e. those sounds were not modulated in

262 frequency), so this parameter was excluded from the comparative analyses. We transformed the

263 overall acoustic dataset and tested it for the distribution fitting. Firstly, continuous variables were

Predano na recenziju u Plos One

264 log10-transformed (DUR, PRR, FM and PF), while discrete (SR and NP) were square root-

265 transformed. We then tested the variables for normal distribution by using Shapiro-Wilks W test

266 with a level of significance P < 0.05. Afterwards, to investigate the sound variation among the

267 Gobius lineage gobies, species were used as grouping variable. Since the assumption of normality

268 was not met, for interspecific comparison we used the non-parametric Kruskal-Wallis H test

269 followed by Dunn’s multiple comparison test (level of significance P < 0.001) to investigate the

270 variation of individual means for each sound variable across species. Individual mean values of

271 sound variables were tested for correlation using the non-parametric Spearman correlation (level

272 of significance P < 0.001) to investigate their mutual relationships. Furthermore, to quantify

273 acoustic variability among the species, we applied multivariate exploratory techniques.

274 Individual means of all sound variables of nine gobiids were compared to test for the overall

275 signal similarity using Principal component analysis (PCA). PCA (in our case based on the

276 correlation matrix) creates a factor space for a set of variables, and therefore we used it

277 specifically to identify the acoustic parameters that explain the most variance among the taxa in

278 the obtained factor space. For the interpretation of PCA results, we used as many factors as the

279 number of eigenvalues > 1.0. In order to discriminate the species according to the acoustic

280 parameters, a forward stepwise Discriminant function analysis (sDFA) was also carried out on

281 the individual mean acoustic variables, with the specific aim to determine which parameters are

282 responsible for species differentiation. In addition, sDFA was also used to assess the probability

283 (classification rate %) at which individual sounds will be classified into the correct taxa.

284 Specifically, sDFA enters variables into the discriminant function model one by one, always

285 choosing the variable that makes the most significant contribution to the discrimination model.

286 Factor structure coefficients were chosen to indicate the correlations between the variables and

287 the discriminant functions. Partial Wilks' Lambda was chosen to indicate the contribution of each

288 variable to the overall discrimination between species. The selection criterion for an acoustic

289 parameter to be entered was F = 1.0, while F = 0.0 (P = 0.01) was the exclusion criteria for removal

Predano na recenziju u Plos One

290 from the analysis. All statistical analyses were performed in STATISTICA® (version 13.6.0., TIBCO,

291 USA) software.

292

293 Authorisations

294 Since all the acoustic data were already published in previous papers, no experimental

295 acoustic work was conducted within the present study. However, all the previous experiments

296 described in this article were compliant with the current laws for animal experimentation in

297 Croatia (Bioethics and Animal Welfare Committee, Faculty of Science, University of Zagreb;

298 permit #251-58-10617-18-14) and with the Venice Declaration (Italy). In addition, the licences

299 525-13/0545-18-2 and 525-1311855-19-2 (Ministry of Agriculture) permitted the field sampling

300 of a Croatian ichthyofauna and licenses released by Regione Veneto (Italy) for scientific fishery of

301 Italian species. As regarding P. nigricans, sampling protocols have been established in compliance

302 with the ethical standards, as approved by the Italian regulations and by local permitting

303 authorities (Umbria Region), who provided the sampling authorizations (Resolution of the

304 Regional Council (DGR) N. 19, session of 16/01/2017). All the experiments were performed in

305 accordance with standard ethological and bioacoustics procedures (avoiding suffering or

306 damaging of fish body parts), meaning that all tested fish, after the laboratory analyses, were

307 returned safely and unharmed to their natural habitat.

308 Results

309 Interspecific acoustic variation and sound properties

310 Perccottus glenii produces thumps sounds, with an irregular waveform and a lack of

311 frequency modulation. However, the nine soniferous gobiids share certain common

312 characteristics of their vocal repertoire, allowing for interspecific comparison (Table S2). All

Predano na recenziju u Plos One

313 acoustic variables differed significantly among the species (Kruskall-Wallis test: SR: H = 46.16, P

314 < 0.001; DUR: H = 60.41, P < 0.001; NP: H = 64.64, P < 0.001; PRR: H = 55.22, P < 0.001; PF: H =

315 48.05, P < 0.001; FM: H = 60.95, P < 0.001; d.f. = 9, n = 73 for each sound property), with at least

316 one species differing from the remaining taxa according to the acoustic variables (Fig 1).

317 Correlation analysis performed on individual mean values of acoustic variables indicated that

318 DUR and NP were mutually and significantly associated (Spearman r = 0.88, n = 67, P < 0.001;

319 Table 2), meaning that as sounds become longer, more pulses are stacked together. On the other

320 hand, the two spectral variables PF and FM were negatively correlated (Spearman r = -0.43, n =

321 67, P < 0.001; Table 2). In the principle component analysis (PCA), the axes PC1 and PC2

322 accounted cumulatively for 63% of the variation, with the first two axes explaining 35.6 and

323 27.6% of the variance, respectively (Table S3). Temporal properties of the sound DUR, NP and

324 PRR were positively associated with PC1, while spectral variables PF (positively) and FM

325 (negatively) contributed to PC2 (Table S3). The PC1 versus PC2 scatterplot of the taxa illustrated

326 the acoustic variation between gobiids according to the call properties (Fig S1).

327

328 Fig 1. Box plot of the six acoustic variables of the sounds produced by ten gobioid species. The

329 midline represents the median, x marks the mean, box values highlights 25 and 75 percentiles

330 while the whiskers indicate minimum and maximum values of the acoustic properties for each

331 species. Abbreviations of the six acoustic variables: SR = sound rate, DUR = sound duration, NP =

332 number of pulses, PRR = pulse repetition rate, PF = peak frequency, FM = frequency modulation.

333

334 Table 2. Spearman correlation coefficient of the relationships between the six acoustic

335 properties. Correlation is based on the individual means of six acoustic properties per species (N

336 = 9).

Variable SR DUR (ms) NP PRR (Hz) PF (Hz) FM (Hz) SR (s/min) 1.000 0.037 0.102 0.035 0.050 0.318

Predano na recenziju u Plos One

DUR (ms) 0.037 1.000 0.886* -0.006 0.178 -0.201 NP 0.102 0.886* 1.000 0.360 0.052 0.109 PRR (Hz) 0.035 -0.006 0.360 1.000 0.115 0.326 PF (Hz) 0.050 0.178 0.052 0.115 1.000 -0.435* FM (Hz) 0.318 -0.201 0.109 0.326 -0.435* 1.000

337 Asterisk (*) marks significant correlation under P < 0.001 criterion.

338

339 Stepwise Discriminant function analysis (sDFA) differentiate the gobies according to their

340 sound properties (Wilks’ Lambda = 0.0002, F28, 264 = 25.28, n = 67, P < 0.001). The first two

341 discriminant functions (DF1 and DF2) cumulatively explained 64.4 and 27.5% of the variation,

342 with the DF1 significantly loaded with call properties PRR and PF, while DF2 showed a positive

343 correlation with NP and DUR (Table 3). In addition, the DFA indicated that individual sounds were

344 correctly classified into corresponding species with an overall 92.5% correct classification rate

345 (Table 4). Accuracy of the classification rate varied among the species, with P. bonelli, G.

346 paganellus, G. niger, P. kessleri and Z. ophiocephalus classified with an accuracy of 100%, while G.

347 cobitis had the lowest fidelity (66.6%), indicating that sounds could be more variable in this

348 species compared to others. Partial Wilks' Lambda (for all variables < 0.5, P < 0.001) indicated

349 that acoustic variables SR, DUR, NP, PRR and PF contributed, in the most significant order, to the

350 overall discrimination contrary to FM which did not contribute (Partial Wilks' Lambda > 0.5, P =

351 0.10). The DFA differentiated several groups of species (Fig 2). Accordingly, P. bonelli, P. kessleri

352 and G. paganellus were clustered on the right side of the diagram, though P. bonelli was slightly

353 more distant along the negative side of DF1 in comparison with the remaining two species. This

354 group was separated from the others due to high PRR and NP values, which contributed

355 significantly to DF1 and DF2. Neogobius fluviatilis, N. melanostomus and P. nigricans were situated

356 on the positive DF1 and negative DF2 axes, mainly because their sounds are characterized by a

357 short duration with low NP and high PRR. Lastly, G. niger, G. cobitis and Z. ophiocephalus are

358 situated on the lower-left part of the diagram, where G. niger, G. cobitis are clustered along the

359 negative DF1 and DF2 sides. Zosterisessor ophiocephalus occupies a separate position from the

Predano na recenziju u Plos One

360 two mentioned Gobius species along the negative DF1 axis (Fig 2). These species produce long

361 sounds with a high NP but low PRR. By extracting the two most important temporal acoustic

362 properties from DFA, DUR and PRR, a scatterplot was built to illustrate the acoustic structure in

363 more details (Fig 3). Species producing short tonal sounds (< DUR, > PRR; N. melanostomus, N.

364 fluviatilis and P. nigricans) clustered on the left-uppermost part of the diagram, while the species

365 situated on the left-lowermost part are characterized by pulsatile sounds (Z. ophiocephalus, G.

366 niger and G. cobitis; intermediate DUR, < PRR). Finally, the three species producing long

367 tonal/complex sounds are positioned in the upper-middle part of the diagram (> DUR, > PRR; P.

368 kessleri, G. paganellus and P. bonelli; see Fig 3).

369

370 Fig 2. Scatterplot of discriminant function 1 (DF1) versus discriminant function 2 (DF2)

371 performed with individual means of the six acoustic properties. Each species, set as a grouping

372 variable, is represented by a different symbol.

373

374 Fig 3. Categorized scatterplot of two temporal variables (DUR versus PRR) highlighting the

375 acoustic variability between the nine gobiid species. For each species, the representative

376 spectrogram in kilohertz is mapped, where brighter colours indicate higher energy intensity.

377 Sounds were recorded at 44.1 kHz and 16-bit resolution while the spectrogram was prepared

378 using AVISOFT software (Hamming window, 512-points FFT, resolution 7 Hz). On the scatterplot,

379 each symbol represents the plot of a selected variable (species mean for DUR) against the value

380 of another selected variable (species mean for PRR) broken down (i.e. categorized) by grouping

381 variable (Species).

382

383 Table 3. Factor structure coefficients from the discriminant function analysis (DFA) representing

384 the correlations between the six acoustic variables and the respective discriminant functions

Predano na recenziju u Plos One

385 (DF). In the DFA, species were set as a grouping variable and individual means of the six acoustic

386 properties as the dependent variables.

Variable DF1 DF2 DF3 DF4 SR (s/min) 0.05 0.06 0.84 0.17 DUR (ms) -0.06 0.52 0.02 -0.79 NP 0.19 0.91 -0.07 -0.33 PRR (Hz) 0.44 0.37 0.02 0.60 PF (Hz) -0.41 0.21 0.22 0.56 FM (Hz) 0.07 -0.05 0.24 0.11 387

388 Table 4. Stepwise classification matrix indicating the percent of cases (individual sounds) that

389 are correctly classified in each group (species) by the classification functions and the cases that

390 are misclassified in each group. Total classification rate is also indicated.

P. P. G. G. G. Z. N. N. P. Taxon % bonelli nigricans paganellus cobitis niger ophiocephalus fluviatilis melanostomus kessleri P. bonelli 100 5 0 0 0 0 0 0 0 0 P. nigricans 75.0 0 3 1 0 0 0 0 0 0 G. paganellus 100 0 0 15 0 0 0 0 0 0 G. cobitis 66.7 0 0 1 4 1 0 0 0 0 G. niger 100 0 0 0 0 5 0 0 0 0 Z. ophiocephalus 100 0 0 0 0 0 8 0 0 0 N. fluviatilis 87.5 0 0 0 0 0 0 7 1 0 N. melanostomus 85.7 0 0 0 0 0 0 1 6 0 P. kessleri 100 0 0 0 0 0 0 0 0 9 Total 92.5 5 3 17 4 6 8 8 7 9

391

392 Phylogenetic affinities between vocal gobiids

393 The molecular analysis of a concatenated dataset, inferred from two nuclear (rag1 & rho)

394 and two mitochondrial (cytb & cox1) molecular markers, allowed us to reconstruct the

395 phylogenetic relationships of nine vocal European gobiids and to build the genetic distance matrix

396 for the pairwise comparison using acoustic data (Table 5). The matrix of 3961 base pairs (bp)

397 contained 30% variable sites, of which 17% are parsimony informative. The sequence lengths of

Predano na recenziju u Plos One

398 individual markers were: cytb 1113, cox1 650, rag1 1458 and rho 740. Phylogenies reconstructed

399 based on the concatenated data using maximum likelihood (ML) and Bayesian inference (BI)

400 method showed identical topologies. Padogobius bonelli is in a sister position to all remaining

401 gobiids (Fig 4). The remaining species formed three well supported topological groups. Two

402 marine gobies, Gobius niger and Zosterisessor ophiocephalus, grouped into one clade, G. cobitis and

403 G. paganellus clustered into the second, while the third group was composed of gobiids

404 distributed in the Ponto-Caspian region (genera Neogobius and Ponticola) and Padogobius

405 nigricans, an Italian freshwater endemic species (Fig 4). In the third group, P. kessleri is a sister

406 taxon to the remaining species, P. nigricans and N. melanostomus cluster as sister species, and

407 Neogobius fluviatilis is a sister taxon in regards to these two species (Fig 4). These results support

408 the monophyly of these vocal Ponto-Caspian species (including Padogobius nigricans) and suggest

409 the polyphyly of the genus Padogobius.

410

411 Fig 4. Bayesian inference phylogenetic relationships between the studied goby species based on

412 concatenated dataset of two mitochondrial (cytb and cox1) and two nuclear markers (rag1 and

413 rho). The numbers on nodes represent posterior probability (BI) and bootstrap support (%, ML)

414 values. Nodes with values ≥0.95 for posterior probability and ≥75% for bootstrap support are

415 considered well supported. For each taxon, a single representative sound waveform was mapped

416 to underline the acoustic affinities between the investigated taxa. Species groups are distinctly

417 shaded: dark grey - Atlantic-Mediterranean gobiids; medium grey - Ponto-Caspian taxa; light grey

418 - odontobutid Perccottus glenii. For P. glenii, two different sounds are mapped in order to follow

419 the evolutionary pattern of acoustic divergence. Sounds were recorded at 44.1 kHz and 16 bit. In

420 addition, for each species, three life-history traits are indicated by ◻ (number of vertebrae: white

421 - less than 28; grey - 27-31; black - more than 28), ◯ (swim bladder: white - absent; black -

422 present) and △ (habitat: white - freshwater; black - marine). Habitat refers to the water type from

423 which individuals for the analysis were captured. Waveforms are not to scale.

Predano na recenziju u Plos One

424

425 Table 5. Genetic distance matrix estimated from the concatenated dataset (cytb, cox1, rag1 and

426 rho) using the p-distance method in MEGA.

G. G. G. Z. P. P. N. N. Taxon cobitis paganellus niger ophiocephalus nigricans bonelli fluviatilis melanostomus G. paganellus 7.73 G. niger 10.20 10.08 Z. ophiocephalus 10.36 9.73 9.35 P. nigricans 9.91 10.44 11.84 10.89 P. bonelli 10.66 11.04 12.29 11.71 11.21 N. fluviatilis 9.40 9.86 10.98 11.06 6.20 10.86 N. melanostomus 9.61 9.94 11.67 11.07 6.05 11.24 5.75 P. kessleri 10.93 11.54 12.52 12.49 10.04 12.04 9.50 9.94

427

428 Acoustic and genetic divergence comparison

429 We performed pairwise comparison between call divergence and genetic diversity to

430 investigate their mutual relatedness in gobiids. Specifically, call divergence matrix was built from

431 Cluster analysis (clustering performed by using six acoustic features), and genetic distance

432 matrix, which was obtained by using the uncorrected p-distance method based on the

433 concatenated data set for all used markers (cytb, cox1, rag1 and rho), and separately for mtDNA

434 (cytb and cox1) and nDNA markers (rag1 and rho). A significant positive correlation was found

435 between acoustic and genetic distance matrices (Mantel test r = 0.470, Z = 2298.756, Ptwo-tailed =

436 0.01; Fig 5), even after performing the matrix permutation test (10,000 repetitions, Ptwo-tailed =

437 0.03) indicating that in vocal gobiids from the Gobius lineage, interspecific divergence in sound

438 follows the same phylogenetic pattern of diversification. In addition, when the acoustic distance

439 was compared with genetic divergence based on nuclear (rag1 and rho) or mitochondrial (cytb

440 and cox1) markers, the results were similar to those with the concatenated dataset. There was a

441 significant positive correlation between acoustic and nuclear distance (Mantel test for nDNA: r

442 =0.485, Z = 793.852, Ptwo-tailed = 0.005; Fig S2), and acoustic and mitochondrial distance (Mantel

Predano na recenziju u Plos One

443 test for mtDNA: r = 0.450, Z = 4195.402, Ptwo-tailed = 0.01; Fig S2), indicating once again that

444 regardless the type of genetic marker and its evolutionary rate, sounds diverge in similar pattern

445 to the phylogenetic affinities in gobiids.

446

447 Fig 5. Correlation between genetic distance and acoustic divergence in nine vocal Gobius lineage

448 gobiids (Mantel test r = 0.47, Pt.t. = 0.01). Genetic distance was estimated from the concatenated

449 dataset (p-distance method), while acoustic distance was estimated from the standardized

450 Manhattan distance metric procedure using species means of the six sound variables. The dashed

451 patter shows linear trend line, while the scatterplot represents the relationship between species

452 genetic differentiation and their acoustic distance.

453

454 Discussion

455 In comparison with the vocalizations of other vertebrates like frogs, birds or mammals,

456 and given the possible absence of the confounding effects of learning [86,87], the relative

457 simplicity and strong high stereotypy of teleost sounds make fish a useful group for studying the

458 evolution of acoustic communication and its association with phylogeny. By correlating acoustic

459 variability with genetic divergence, we sought to elucidate whether sounds in vocal gobiids

460 (Gobius lineage) have a phylogenetic basis, and did stochastic evolutionary forces play a

461 prominent role in signal divergence. No similar investigation has never been performed in gobiids

462 or any other teleost group to date, and our results could be meaningful since speciation may be

463 accelerated by the separation of signalling systems [20]. Since various communication signals,

464 including acoustic signals reveal the identity of the signalling animal, they may be involved in

465 species diversification [11].

Predano na recenziju u Plos One

466 This acoustic analysis allowed us to discriminate between nine gobiids according to the

467 spectro-temporal properties of their sounds, indicating that each taxon produces species-specific

468 calls characterized by a unique set of variables allowing for interspecific differentiation. Since

469 most gobies live in natural sympatry with at least one other species [58-61], we believe that these

470 sounds could act as accurate species-discrimination traits. From the acoustic analyses including

471 PCA and sDFA, certain sound properties, especially temporal NP, DUR and PRR and spectral PF

472 appear to be responsible for the observed interspecific call divergence. Since these sound

473 properties accounted for most of the variation among species in both PCA and sDFA (variable-

474 factor correlations ranged from 0.4 to 0.9), they can be considered the main acoustic components

475 carrying the phylogenetic signal. However, since in the acoustic analyses, NP was one of the

476 important sound features, but it was strongly correlated with DUR (Spearman r > 0.8), DUR can

477 be considered an independent and phylogenetically more informative trait than NP. This is

478 corroborated by the overall similarity in taxa composition comparing the sDFA scatterplot with

479 the DUR versus PRR diagram, where both highlighted a similar pattern of acoustic divergence

480 among the studied gobiids. Moreover, the sDFA emphasized that the individual sounds were

481 accurately attributed to corresponding species with high classification fidelity (> 90%), meaning

482 acoustic signals could indeed reflect phylogenetic taxon affiliation. The first comparative study

483 on acoustic signals produced by Mediterranean gobies [49] proposed that relationships between

484 Mediterranean gobies could be inferred using the signal structure as a reliable indicator of taxon

485 affiliation, given the strong relationships between acoustic affinities and species traits. However,

486 without genetic data, this was a long-standing hypothesis corroborated by this study.

487 Our phylogenetic analysis strongly indicated that P. bonelli is separated from the rest of

488 the investigated taxa, occupying an isolated position on both the sDFA phenogram and

489 concatenated phylogenetic tree. This is in agreement with the known vocal diversity of P. bonelli,

490 with three different sound types reported to date [49,80]. Among the Atlantic-Mediterranean

491 gobiids (Gobius and Zosterisessor genera), deeper phylogenetic relationships remain unresolved,

492 though some interesting observations can be drawn from our results. For Z. ophiocephalus and G.

Predano na recenziju u Plos One

493 niger, an isolated group on the genetic tree, the observed phylogenetic relationship coincides with

494 their similar acoustic structure of pulsatile sounds and substrate preferences [mud or silt; 60].

495 Likewise, their close phylogenetic affiliation has been confirmed using DNA sequences from both

496 mitochondrial and nuclear markers [42,43,45], although these studies included a smaller number

497 of species for phylogenetic analysis. In the PCA and sDFA, both species occupied the same acoustic

498 space, even though Z. ophiocephalus had a slightly more isolated position in both PCA and sDFA

499 scatterplot, due likely to the higher spectral sound properties, mostly PF. The observed acoustic

500 and genetic relationship between Z. ophiocephalus and G. niger sheds light on the taxonomic

501 position of Z. ophiocephalus, as some authors have suggested a close phylogenetic relationship

502 with the genus Gobius. On the other hand, G. cobitis and G. paganellus are mostly found on rocky

503 bottoms, and they share certain phenotypic traits [e.g. colouration pattern and sagittal otolith

504 shape; 60,88]. The species inhabiting the Ponto-Caspian region (genera Neogobius and Ponticola),

505 produce only one sound type [34,50], justifying their clustering into one well-supported clade

506 [this study and 42,43,46,89]. However, from our results, a certain degree of acoustic variability is

507 evident between the Neogobius group (N. fluviatilis, N. melanostomus and Padogobius nigricans),

508 characterized by the production of short tonal sounds, and Ponticola kessleri, which was

509 separated from the remaining Ponto-Caspian species in the PCA and sDFA diagrams, likely due to

510 its long, frequency-modulated sounds, similar to G. paganellus vocalizations. This is interesting

511 since both species (together with P. bonelli) share similar bottom preferences (rocks or coarse

512 gravel), although P. kessleri is a freshwater resident. This observation might suggest that tonal

513 sounds are more suitable for hard-bottom transmission, although this should be examined in

514 future studies by investigating the ecological adaptations of gobies to certain habitat conditions.

515 Recent studies have suggested that rocky or pebbly substrates, inhabited by the bottom-dwelling

516 gobies producing tonal sounds (like G. paganellus, P. kessleri and P. bonelli), are unfavourable for

517 sound emission, due to the low-frequency ambient noise and short-range transmission of sounds

518 [90,91]. [92] proposed that tonal sounds could possess characteristics enabling longer-range

519 transmission than pulsatile sounds, since the acoustic structure is simpler. In addition, according

Predano na recenziju u Plos One

520 to [92], waveform differences between pressure and particle velocity spectra are less expressed

521 in tonal than pulsatile sounds. In vertebrates, frequency modulated sounds (like the tonal sounds

522 of gobiids) are generally long-range signals [93]. Furthermore, the close affinity of P. nigricans, an

523 Italian endemic goby, with the Ponto-Caspian group in both the sDFA and the phylogenetic tree

524 is phylogenetically interesting. According to [58], P. nigricans is similar to the Ponto-Caspian

525 gobies due to the larger number of vertebrae (> 29), and the presence of a reduced swimming

526 bladder and the presence of head canals, while P. bonelli has a swim bladder but lacks head canals

527 [58]. Likewise, recent acoustic studies indicated that Padogobius could be polyphyletic [34,50],

528 supporting previous molecular findings that emphasized that the two Padogobius species are of

529 independent origin [94-96]. However, no studies have used robust sound data set to investigate

530 acoustic diversity and to combined genetic data with sound variability to empirically confirm this

531 hypothesis. Our comparative study strongly corroborated these hypotheses, indicating that the

532 genus Padogobius is truly polyphyletic. Sounds, as shown here, proved to be a valuable species-

533 specific trait in the European gobiids, and a suitable basis for future phylogenetic studies.

534 In young or emerging species, acoustic signals may serve as isolating mechanisms, leading

535 to intraspecific acoustic variability. It has long been debated whether selection or drift have

536 relative importance in the process of speciation [97,98]. To demonstrate that stochastic (also

537 called ‘neutral’), and not deterministic (‘adaptive’) evolution is most important in driving acoustic

538 differences, divergence in acoustic traits should be empirically confirmed to increase linearly

539 with genetic distance, with little or no effect of selection [11,20]. Although we did not test the

540 effects of selection, a positive linear correlation between acoustic distance and genetic divergence

541 for all investigated species was obtained, using both a concatenated molecular dataset and

542 individual mtDNA/nDNA markers. This indicates that sounds diverge in the same manner as DNA

543 sequences, regardless the marker type and its evolutionary rate. In the evolutionary sense, these

544 findings suggest that stochastic forces could be more responsible for shaping acoustic divergence

545 than deterministic processes in European gobiids. In the presence of a positive correlation, most

546 studies have emphasized that drift is the main driver behind differentiation in calls between

Predano na recenziju u Plos One

547 species [11,99,100]. However, we propose that this should be verified by future studies exploring

548 the intensity of sexual selection between closely related species. In addition, other factors such as

549 social learning [101] and mutation-order processes [102] could drive acoustic differentiation. In

550 learned vocal signals [e.g., bird song, 103; whale song, 104], cultural transmission and copying

551 errors are major drivers of stochastic divergence. The ability of gobiids to learn sounds has never

552 been investigated. Considering the recent findings of other teleosts, suggesting that the sounds in

553 fish are innate [86,87], the effects of social selection as a driving force for the observed divergence

554 can be excluded at this time. Mutation-order processes over time can cause the linear

555 accumulation of acoustic differences [105,106], resulting in a highly complex interaction between

556 such processes and drift. However, some gobiids hybridize [107], violating reproductive isolation

557 as one of the main criteria for mutation-order speciation. Moreover, we have no precise

558 knowledge whether acoustic divergence is associated with ecological or sexual trait divergence,

559 and therefore their indirect influence on overall divergence cannot be excluded, since this

560 divergence may result from a combination of selection and drift [20,108]. Geographic variation

561 or ecological selection in gobiid vocalizations is still largely unexplored, with few reports of

562 “dialects” in gobiids [37,50]. Therefore, comparing genetic distance and sound variability shaped

563 by regional or habitat selection was not feasible in the present study [13,25,29]. We suggest that

564 neutral evolution (i.e. stochastic processes) drives acoustic signal divergence in gobiids, although

565 the influence of selection should not be neglected, since it likely has some impact on overall

566 diversification. Sounds, including their acoustic features, carry important phylogenetic signals for

567 species recognition, although deterministic processes such as sexual selection or habitat

568 adaptations could have an indirect role in acoustic divergence. The exact degree of the

569 phylogenetic signal carried by the acoustic features of gobiid sounds and their rate of evolution

570 remains unclear. In their elaborate study, [106] tested whether the differences in male sexual

571 signal (nuptial colour) were correlated with environmental, genetic or geographic distances in

572 darters. From the observed correlation between overall male colour differences (i.e. scores for

573 discrete colour categories) and genetic divergence, authors concluded that a single phenotypic

Predano na recenziju u Plos One

574 trait, i.e. breeding coloration of the males (or in our case the sound), could possibly be a

575 combination of various independent (continuous and discrete) characters, each operating under

576 different selective regime. Having this in mind, we can expect that certain sound properties (as

577 previously mentioned NP, DUR, PRR and PF) would carry different levels of phylogenetic

578 information. The lower phylogenetic signal of the acoustic traits than morphological traits

579 showed that sexual selection could be an important driver of diversification in bats and glass frogs

580 [18,19]. If sexual selection was the main driver of acoustic divergence in gobiids, we would not

581 expect a positive association between song similarity and genetic divergence, since acoustic

582 signals would then diverge faster than genetic loci [11]. It is interesting that sympatric species in

583 this study (i.e. gobies occurring together in the same region, but preferring different

584 microhabitats; Z. ophiocephalus with G. niger; G. paganllus with G. cobitis; P. kessleri with N.

585 melanostomus and N. fluviatilis) differed acoustically, which may be to avoid signal interference

586 [109], since sympatric taxa compete for acoustic niches.

587 From our results and the existing literature, we were able to build the acoustic hypothesis

588 explaining the evolution of sound communication in soniferous gobioids (Fig 6). Briefly, short

589 tonal sound and thumps (i.e. short irregular pulses) noted in P. glenii (Odontobutidae) represent

590 stem signals, i.e. the symplesiomorphic condition for gobioids. Observed phylogenetic pattern of

591 sound production in gobioids suggests that acoustic signals probably evolved from multiple brief,

592 rapidly repeated sounds (like thumps and tonal) to long and frequency modulated calls (pulsatile

593 and complex) documented in vocal Gobius lineage gobiids. In these species (P. bonelli and genus

594 Gobius), stem sound types gave the structural elements for the construction of complex calls,

595 while the Neogobius group returned to the ancestral acoustic state or shares homoplasious trait

596 with odontobutids (i.e. production of short tonal sounds).

597 Fig 6. Diagram depicting the divergence of acoustic signals between soniferous gobioids

598 following the evolutionary hypothesis. Affinities between the gobiids follow the interspecific

599 relationships obtained from the concatenated Bayesian inference (BI) phylogenetic tree from the

Predano na recenziju u Plos One

600 present study. Abbreviations: DUR = duration (ms), PRR = pulse repetition rate, FM = frequency

601 modulation (Hz), s., sounds.

602

603 Conclusion

604 In summary, our results corroborate previous evidence that interspecific acoustic signals

605 are highly diverse among vocal European gobiids. Each species from our study is recognisable

606 based on its sound structure and spectro-temporal features, making its vocalisations a highly

607 species-specific trait. Since P. nigricans clustered in the same acoustic and genetic topology with

608 Neogobius spp., we suggest that the genus Padogobius is polyphyletic and that P. nigricans is

609 closely related to Ponto-Caspian gobies. This result is further supported by the morphological

610 traits, i.e. vertebrate number and absence of a swimbladder. Furthermore, our comparative

611 acoustic-genetic analyses suggest that signal divergence in European gobiids was likely caused

612 by stochastic processes that shaped acoustic diversity. This is corroborated by the observed

613 pattern of sound divergence, which correlated linearly with genetic distance. Therefore, we

614 propose that certain acoustic properties of gobiid sounds carry a phylogenetic signal responsible

615 for species recognition. Future studies should quantify the exact level of phylogenetic information

616 present in these acoustic properties. However, these studies should not neglect the influence of

617 deterministic process in the ecological and geographical divergence in gobiid sounds, and their

618 roles in acoustic divergence should be determined. Phonotaxic experiments are necessary to test

619 the influence of sexual selection in the sound differentiation process, highlighting sexually-

620 selected acoustic traits. In conclusion, we strongly suggest that sounds in European gobiids are

621 species-specific and evolutionarily based, presenting a promising phylogenetic tool for future

622 comparative studies aiming to resolve their affinities and taxonomic status.

623

Predano na recenziju u Plos One

624 Acknowledgements

625 The authors are grateful to technician S. Vajdić and R. Karlović for their dedicated help during fish

626 sampling. M. Kovačić provided the genetic material for the two marine species (G. cobitis and G.

627 paganellus), while M. Smederevac-Lalić and S. Skorić provided the live specimens of Perccottus

628 glenii. Finally, we thank I. Bacelj for the help with statistical issues and L. Zanella for proofreading

629 the paper.

630

631 Author contributions

632 Conceptualization: Sven Horvatić, Stefano Malavasi

633 Data curation: Sven Horvatić

634 Formal analysis: Sven Horvatić, Jasna Vukić, Ivana Buj, Lucija Raguž, Lucija Ivić, Stefano Malavasi

635 Funding acquisition: Jasna Vukić, Davor Zanella

636 Investigation: Sven Horvatić, Jasna Vukić, Stefano Malavasi

637 Methodology: Sven Horvatić, Stefano Malavasi, Davor Zanella, Jasna Vukić, Radek Šanda, Zoran

638 Marčić, Marko Ćaleta, Perica Mustafić, Francesco Cavraro, Massimo Lorenzoni

639 Project administration: Sven Horvatić, Davor Zanella, Jasna Vukić

640 Software: Sven Horvatić, Jasna Vukić

641 Supervision: Stefano Malavasi, Davor Zanella

642 Validation: Sven Horvatić, Stefano Malavasi, Jasna Vukić,

643 Visualization: Sven Horvatić, Perica Mustafić

644 Writing – original draft: Sven Horvatić, Jasna Vukić

Predano na recenziju u Plos One

645 Writing – review & editing: Sven Horvatić, Stefano Malavasi, Jasna Vukić, Radek Šanda, Zoran

646 Marčić, Marko Ćaleta, Massimo Lorenzoni, Perica Mustafić, Ivana Buj, Lucija Raguž, Lucija Ivić,

647 Francesco Cavraro & Davor Zanella

648

649 References

650 1. Simmons AM. Perspectives and Progress in Animal Acoustic Communication. In: Simmons A,

651 Fay RR, editors. Acoustic Communication. Springer; 2003. pp. 1-4.

652 2. Ladich F, Winkler H. Acoustic communication in terrestrial and aquatic vertebrates. J Exp Biol.

653 2017;220: 2306-2317.

654 3. Crawford JD, Cook AP, Heberlein AS. Bioacoustic behaviour of African fishes (Mormyridae):

655 potential cues for species and individual recognition in Pollymirus. J Acoust Soc Am. 1997;102:

656 1200-1212.

657 4. Lobel PS. Possible species specific courtship sounds by two sympatric cichlids fishes in Lake

658 Malati, Africa. Environ Biol Fishes. 1998;52: 443-452.

659 5. Amorim MCP, Vasconcelos RO, Fonseca PJ. Fish sounds and mate choice. In: Ladich F, editor.

660 Sound Communication in Fishes. Springer; 2015. pp. 1-35.

661 6. Pedroso S, Barber I, Svensson O, Fonseca PJ, Amorim MC. Courtship sounds advertise species

662 identity and male quality in sympatric Pomatoschistus spp. gobies. PLoS One. 2013. doi:

663 10.1371/journal.pone.0064620.

664 7. Blom EL, Mück I, Heubel K, Svensson O. Acoustic and visual courtship traits in two sympatric

665 marine Gobiidae species - Pomatoschistus microps and Pomatoschistus minutus. Environ Biol

666 Fishes. 2016;99: 999-1007.

667 8. Zeyl JN, Malavasi S, Holt DE, Noel P, Lugli M, Johnston CE. Convergent aspects of acoustic

668 communication in darters, sculpins and gobies. In: Sisneros AJ editor. Fish Hearing and

669 Bioacoustics: An Anthology in Honor of Arthur N. Popper and Richard R. Fay. Springer; 2016.

670 pp. 93-120.

Predano na recenziju u Plos One

671 9. Ritchie MG. Sexual selection and speciation. Annu Rev Ecol Evol Syst. 2007; 38: 79-102.

672 10. Seddon N, Botero CA, Tobias JA, Dunn PO, MacGregor HEA, Rubenstein DR, et al. Sexual

673 selection accelerates signal evolution during speciation in birds. Proc. Biol. Sci. 2013. doi:

674 10.1098/rspb.2013.1065.

675 11. Wilkins M, Seddon N, Safran R. Evolutionary divergence in acoustic signals: Causes and

676 consequences. Trends Ecol Evol. 2012;28: 156-166.

677 12. Mendelson TC, Martin MD, Flaxman SM. Mutation-order divergence by sexual selection:

678 diversification of sexual signals in similar environments as a first step in speciation. Ecol. Lett.

679 2014;17: 1053-1066.

680 13. Sun K, Luo L, Kimball RT, Wei X, Jin L, Jiang T, et al. Geographic variation in the acoustic traits

681 of Greater Horseshoe bats: testing the importance of drift and ecological selection in

682 evolutionary processes. PloS One. 2013. doi: 10.1371/journal.pone.0070368.

683 14. Amorim MCP, Vasconcelos RO, Bolgan M, Pedroso SS, Fonseca PJ. Acoustic communication in

684 marine shallow waters: Testing the acoustic adaptive hypothesis in sand gobies. J Exp Biol.

685 2018. doi: 10.1242/jeb.183681.

686 15. Grzywacz B, Heller KG, Warchałowska-Śliwa E, Karamysheva T, Chobanov D. Evolution and

687 systematics of Green Bush-crickets (Orthoptera: Tettigoniidae: Tettigonia) in the Western

688 Palaearctic: testing concordance between molecular, acoustic, and morphological data. Org

689 Divers Evol. 2017;17: 213-228.

690 16. Mélotte G, Raick X, Regis V, Parmentier E. Origin and evolution of sound production in

691 Serrasalmidae. Biol J Linn Soc. 2019;128: 403-414.

692 17. Boul KE, Funk WC, Darst CR, Cannatella DC, Ryan MJ. Sexual selection drives speciation in an

693 Amazonian frog. Proc R Soc B: Biol Sci. 2007;274: 399-406.

694 18. Puechmaille SJ, Borissov IM, Zsebok S, Allegrini B, Hizem M, Kuenzel S, et al. Female mate

695 choice can drive the evolution of high frequency echolocation in bats: a case study with

696 Rhinolophus mehelyi. PloS One. 2014. doi:10.1371/journal.pone.0103452.

Predano na recenziju u Plos One

697 19. Sulbarán MDE, Simões P, Gonzalez-Voyer A, Castroviejo-Fisher S. Neotropical frogs and

698 mating songs: The evolution of advertisement calls in glassfrogs. J Evol Biol. 2018;32: 163-

699 176.

700 20. Irwin D, Thimgan M, Irwin J. Call divergence is correlated with geographic and genetic

701 distance in Greenish Warblers (Phylloscopus trochiloides): a strong role for stochasticity in

702 signal evolution? J Evol Biol. 2008;21: 435-48.

703 21. Jones G. Acoustic signals and speciation: the roles of natural and sexual selection in the

704 evolution of cryptic species. Adv Study Behav. 1997;26: 317-354.

705 22. Mendelson TC, Shaw KL. Sexual behaviour: rapid speciation in an arthropod. Nature.

706 2005;433: 375-376.

707 23. Clegg SM , Degnan SM, Moritz C, Estoup A, Kikkawa J, Owens IPF. Microevolution in island

708 forms: the roles of drift and directional selection in morphological divergence of a passerine

709 bird. Evolution. 2002;56: 2090-2099.

710 24. Uyeda JC, Arnold SJ, Hohenlohe PA, Mead LS. Drift promotes speciation by sexual selection.

711 Evolution. 2009;63: 583-594.

712 25. Lee KH, Shaner PJ, Lin YP, Lin SM. Geographic variation in advertisement calls of a Microhylid

713 frog - testing the role of drift and ecology. Ecol Evol. 2016;6: 3289-3298.

714 26. Päckert M, Martens J, Sun Y, Veith M. The radiation of the Seicercus burkii complex and its

715 congeners (Aves: Sylviidae): molecular genetics and bioacoustics. Org Divers Evol. 2004;4:

716 341-364.

717 27. Percy D, Taylor G, Kennedy M. Psyllid communication: acoustic diversity, mate recognition

718 and phylogenetic signal. INVERTEBR SYST. 2006;20: 431-445.

719 28. Toews D, Irwin D. Cryptic speciation in a Holarctic passerine revealed by genetic and

720 bioacoustic analyses. Mol Ecol. 2008;17: 2691-705.

721 29. Thinh VN, Hallam C, Roos C, Hammerschmidt K. Concordance between vocal and genetic

722 diversity in crested gibbons. BMC Evol Biol. 2011. doi: 10.1186/1471-2148-11-36.

Predano na recenziju u Plos One

723 30. Velásquez JN, Marambio J, Brunetti E, Méndez MA, Vásquez RA, Penna M. Bioacoustic and

724 genetic divergence in a frog with a wide geographical distribution. Biol J Linn Soc. 2009;110:

725 142-155.

726 31. MacDougall-Shackleton E, MacDougall-Shackleton S. Cultural and genetic evolution in

727 mountain whitecrowned sparrows: song dialects are associated with population structure.

728 Evolution. 2001. doi: 10.1111/j.0014-3820.2001.tb00769.x.

729 32. Christiansons S, Swallow J, Wilkinson G. Rapid evolution of postzygotic reproductive isolation

730 in stalk-eyed flies. Evolution. 2005;59, 849-857.

731 33. Amorim MCP, Neves, ASM. Acoustic signalling during courtship in the painted goby,

732 Pomatoschistus pictus. J Mar Biol Assoc UK. 2007;87: 1017-1023.

733 34. Horvatić S, Cavraro F, Zanella D, Malavasi S. Sound production in the Ponto-Caspian goby

734 Neogobius fluviatilis and acoustic affinities within the Gobius lineage: implications for

735 phylogeny. Biol J Linn Soc. 12016;17: 564-573.

736 35. Amorim MCP, Pedroso SS, Bolgan M, Jordão JM, Caiano M, Fonseca PJ. Painted gobies sing their

737 quality out loud: Acoustic rather than visual signals advertise male quality and contribute to

738 mating success. Funct Ecol. 2012;27: 289-298.

739 36. Parmentier E, Petrinisec M, Fonseca PJ, Amorim MCP. Sound-production mechanism in

740 Pomatoschistus pictus. J Exp Biol. 2017;220: 4374-4376.

741 37. Parmentier E, Kéver L, Boyle K, Corbisier Y, Sawelew L, Malavasi S. Sound production

742 mechanism in Gobius paganellus (Gobiidae). J Exp Biol. 2013;216: 3189-3199.

743 38. Nelson JS, Grande TC, Wilson MV. Fishes of the World. 5th ed. Wiley; 2016. pp. 1-12.

744 39. Kuang T, Tornabene L, Li J, Jiang J, Chakrabarty P, Sparks JS. Phylogenomic analysis on the

745 exceptionally diverse fish clade Gobioidei (Actinopterygii: Gobiiformes) and data-filtering

746 based on molecular clocklikeness. Mol Phylogenetics Evol. 2018;128: 192-202.

747 40. McCraney WT, Thacker CE, Alfaro ME. Supermatrix phylogeny resolves goby lineages and

748 reveals unstable root of Gobiaria. Mol Phylogenetics Evol. 2020;151: 106862.

Predano na recenziju u Plos One

749 41. Thacker CE. Phylogeny of Gobioidei and placement within Acantomorpha with a new

750 classification and investigation of diversification and character evolution. Copeia. 2009;2009:

751 93-104.

752 42. Agorreta A, San Mauro D, Schliewen U, Van Tassell JL, Kovačić M, Zardoya R, Rüber L.

753 Molecular phylogenetics of Gobioidei and phylogenetic placement of European gobies. Mol

754 Phylogenet Evol. 2013;69: 619-633.

755 43. Thacker CE, Roje DM. Phylogeny of Gobiidae and identification of gobiid lineages. Syst

756 Biodivers. 2011;9: 329-347.

757 44. Kovačić M, Šanda R. A new species of Gobius (Perciformes: Gobiidae) from the

758 Mediterranean Sea and the redescription of Gobius bucchichi. J Fish Biol. 2016;88: 1104-

759 1124.

760 45. Iglésias SP, Vukić J, Sellos DY, Soukupová T, Šanda R. Gobius xoriguer, a new ofshore

761 Mediterranean goby (Gobiidae), and phylogenetic relationships within the genus Gobius.

762 Ichthyol Res. 2020. doi: 10.1007/s10228-020-00797-9.

763 46. Neilson ME, Stepien CA. Escape from the Ponto-Caspian: evolution and biogeography of an

764 endemic goby species flock (Benthophilinae: Gobiidae: Teleostei). Mol Phylogenet Evol.

765 2009;52: 84-102.

766 47. Takemura A. Acoustical behaviour of the freshwater goby Odontobutis obscura. Bull Jpn Soc

767 Sci Fish. 1984;50: 561– 564.

768 48. Horvatić S, Malavasi S, Parmentier E, Marčić Z, Buj I, Mustafić P, et al. Acoustic

769 communication during reproduction in the basal gobioid Amur sleeper and the putative

770 sound production mechanism. J Zool. 2019;309: 269-279.

771 49. Malavasi S, Collatuzzo S, Torricelli P. Interspecific variation of acoustic signals in

772 Mediterranean gobies (Perciformes, Gobiidae): comparative analysis and evolutionary

773 outlook. Biol J Linn Soc. 2008;93: 763-778.

774 50. Horvatić S, Bem L, Malavasi S, Marčić Z, Buj I, Mustafić P, et al. Comparative analysis of sound

775 production between the bighead goby Ponticola kessleri and the round goby Neogobius

Predano na recenziju u Plos One

776 melanostomus: Implications for phylogeny and systematics. Environ Biol Fishes. 2019:102:

777 727-739.

778 51. Amorim MCP, Knight ME, Stratoudakis Y, Turner GF. Differences in sounds made by courting

779 males of three closely related Lake Malawi cichlid species. J Fish Biol. 2004;65: 1358-1371.

780 52. Amorim MCP, Vasconcelos RO. Variability in the mating calls of the Lusitanian toadfish

781 Halobatrachus didactylus: cues for potential individual recognition J Fish Biol. 2008;73: 1267-

782 1283.

783 53. Rice AN, Bass, AH. Novel vocal repertoire and paired swimbladders of the three-spined

784 toadfish, Batrachomoeus trispinosus: insights into the diversity of the Batrachoididae. J Exp

785 Biol. 2009;212: 1377-1391.

786 54. Parmentier E, Lecchini D, Frederich B, Brié C, Mann D. Sound production in four damselfish

787 (Dascyllus) species: Phyletic relationships. Biol J Linn Soc. 2009;97: 928-940.

788 55. Melotte G, Vigouroux R, Michel C, Parmentier E. Interspecific variation of warning calls in

789 piranhas: a comparative analysis. Sci Rep. 2016. doi: 10.1038/srep36127.

790 56. Bolgana M, Crucianelli A, Mylonas CC, Henry S, Falguière JC, Parmentier E. Calling activity and

791 calls' temporal features inform about fish reproductive condition and spawning in three

792 cultured Sciaenidae species. Aquaculture. 2020;524: 735243.

793 57. Miller PJ. Gobiidae I. The Freshwater Fishes of Europe. Wiesbaden: Aula-Verlag; 2003. pp.

794 157-398.

795 58. Miller, P.J. Gobiidae II. The Freshwater Fishes of Europe. In: Miller PJ, Economidis PS, editors.

796 Gobiidae II. Wiesbaden: Aula-Verlag; 2004. pp. 157-398.

797 59. Kovačić M, Patzner RA. North-Eastern Atlantic and Mediterranean gobies. In: Patzner RA, Van

798 Tassell JL, Kovačić M, Kapoor BG, editors. The Biology of Gobies. Science Publishers, CRC

799 Press; 2011. pp. 177-193.

800 60. Jardas I. Jadranska Ihtiofauna. Školska knjiga, Zagreb; 1996. pp. 331-359.

Predano na recenziju u Plos One

801 61. Freyhof J. Diversity and Distribution of Freshwater Gobies from the Mediterranean, the Black

802 and Caspian Seas. In: Patzner RA, Van Tassell JL, Kovačić M, Kapoor BG, editors. The Biology

803 of Gobies. Science Publishers, CRC Press; 2011. pp. 279-288.

804 62. Simonovic PD. Phylogenetic relationships of Ponto‐Caspian gobies and their relationship to

805 the Atlantic—Mediterranean Gobiinae. J Fish Biol. 1999;54: 533-555.

806 63. Šanda R, Choleva JL, Šedivá JA, Humka S, Wilson IF. Distribution of loach fishes (Cobitidae,

807 Nemacheilidae) in Albania, with genetic analysis of populations of Cobitis ohridana. Folia Zool.

808 2008;57: 42-50.

809 64. Akihito, Iwata A, Kobayashi T, Ikeo K, Imanishi T, Ono H, et al. Evolutionary aspects of gobioid

810 fishes based upon a phylogenetic analysis of mitochondrial cytochrome b genes. Gene.

811 2000;259: 5-15.

812 65. Machordom A, Doadrio, I. Evidence of a Cenozoic Betic-Kabilian connection based on

813 freshwater fish phylogeography (Luciobarbus, Cyprinidae). Mol Phylogenet Evol. 2001;18:

814 252-263.

815 66. López A, Chen W, Ortí G. Esociform phylogeny. Copeia. 2004;2004: 449-464.

816 67. Ward RD, Zemlak TS, Innes BH, Last PR, Hebert PDN. DNA barcoding Australia´s fish species.

817 Philos Trans R Soc Lond B Biol Sci. 2005;360: 1847-1857.

818 68. Geiger MF , Herder F, Monaghan MT, Almada V, Barbieri R, Bariche M, et al. Spatial

819 heterogeneity in the Mediterranean Biodiversity Hotspot affects barcoding accuracy of its

820 freshwater fishes. Mol Ecol Resour. 2014;14: 1210-1221.

821 69. Taylor MS, Hellberg ME. Marine radiations at small geographic scales: speciation in

822 Neotropical reef gobies (Elacatinus). Evolution. 2005;59: 374-385.

823 70. Neilson ME, Stepien CA. Evolution and phylogeography of the tubenose goby genus

824 Proterorhinus (Gobiidae: Teleostei): evidence for new cryptic species. Biol J Linn Soc. 2009;96:

825 664-684.

826 71. Hall TA. BioEdit: a user-friendly biological sequence alignment editor and analysis program

827 for Windows 95/98/NT. Nucleic Acids Symposium Series. 199;41: 95-98.

Predano na recenziju u Plos One

828 72. Simmons MP, Gatesy J. Coalescence vs. concatenation: Sophisticated analyses vs. first

829 principles applied to rooting the angiosperms. Mol Phylogenet Evol. 2012;91: 98-122.

830 73. Tonini J, Moore A, Stern D, Shcheglovitova M, Ortí G. Concatenation and species tree methods

831 exhibit statistically indistinguishable accuracy under a range of simulated conditions. PLOS

832 Currents Tree of Life. 2015. doi:

833 10.1371/currents.tol.34260cc27551a527b124ec5f6334b6be.

834 74. Lanfear R, Frandsen PB, Wright AM, Senfeld T, Calcott B. PartitionFinder 2: New methods for

835 selecting partitioned models of evolution for molecular and morphological phylogenetic

836 analyses. Mol Biol Evol. 2016;34: 772-773.

837 75. Ronquist F , Teslenko M, van der Mark P, L Ayres DL, Darling A, Höhna S, et al. MrBayes 3.2:

838 efficient Bayesian phylogenetic inference and model choice across a large model space. Syst

839 Biol. 2012;61: 539-542.

840 76. Stamatakis A. RAxML version 8: a tool for phylogenetic analysis and post-analysis of large

841 phylogenies. Bioinformatics. 2014;30: 1312‐1313.

842 77. Miller MA, Schwartz T, Pickett BE, He S, Klem EB, Scheuermann RH, et al. A RESTful API for

843 Access to Phylogenetic Tools via the CIPRES Science Gateway. Evol Bioinf. 2015;11: 43-48.

844 78. Tamura K, Stecher G, Peterson D, Filipski A, Kumar S. MEGA6: molecular evolutionary genetics

845 analysis version 6.0. Mol Biol Evol. 2013;30: 2725-2729.

846 79. Swofford DL. PAUP*. Phylogenetic Analysis Using ParFIny (*and other methods). Version

847 4.0b10a. Sinauer Associates, Sunderland; 2002.

848 80. Lugli M, Torricelli P, Pavan G, Mainardi D. Sound production during courtship and spawning

849 among freshwater gobiids (Pisces, Gobiidae). Mar Freshwater Behav Physiol. 1997;29: 109-

850 126.

851 81. Lugli M, Pavan G, Torricelli P, Bobbio L. Spawning vocalizations in male freshwater gobiids

852 (Pisces, Gobiidae). Environ Biol Fish. 1995;43: 219-231.

Predano na recenziju u Plos One

853 82. Lugli M, Torricelli P, Pavan G, Miller PJ. Breeding sounds of male Padogobius nigricans with

854 suggestions for further evolutionary study of vocal behaviour in gobioid fishes. J Fish Biol.

855 1996;49: 648-657.

856 83. Mantel N. The detection of disease clustering and a generalized regression approach. Cancer

857 Research. 1967;27: 209-220.

858 84. Mantel N, Valand RS. A technique of nonparametric multivariate analysis. Biometrics.

859 1970;26: 547-558.

860 85. Rosenberg MS, Anderson CD. PASSaGE: pattern analysis, spatial statistics and geographic

861 exegesis. Version 2. Methods Ecol Evol. 2011;2: 229-232.

862 86. Johnston CE, Buchanan HM. Learned or innate production of acoustic signals in fishes: a test

863 using a cyprinid. Environ Biol Fish. 2007;78: 183-187.

864 87. Longrie N, Fine ML, Parmentier E. Innate sound production in the cichlid Oreochromis

865 niloticus. J Zool. 2008;276: 266-275.

866 88. Lombarte A, Miletić M, Kovačić M, Otero-Ferrer JL, Tuset VM. Identifying sagittal otoliths of

867 Mediterranean Sea gobies: variability among phylogenetic lineages. J Fish Biol. 2018;92:

868 1768-1787.

869 89. Tornabene L, Chen Y, Pezold F. Gobies are deeply divided: phylogenetic evidence from nuclear

870 DNA (Teleostei: Gobioidei: Gobiidae). Syst Biodivers. 2013;11: 1-17.

871 90. Lugli M, Yan HY, Fine M. Acoustic communication in two freshwater gobies: The relationship

872 between ambient noise, hearing thresholds and sound spectrum. J Comp Physiol (A).

873 2003;189: 309-320.

874 91. Lugli M, Fine ML. Acoustic communication in two freshwater gobies: ambient noise and short-

875 range propagation in shallow streams. J Acoust Soc Am. 2003;114: 512-521.

876 92. Lugli M, Fine ML. Stream ambient noise, spectrum and propagation of sounds in the goby

877 Padogobius martensii: Sound pressure and particle velocity. J Acoust Soc Am. 2007;122: 2881-

878 2892.

Predano na recenziju u Plos One

879 93. Wiley RH, Richards DG. Adaptations of acoustic communication in birds: sound transmission

880 and signal detection. In: Kroodsma DE, Miller EH. Acoustic Communication in Birds. Academic

881 Press; 1982. pp. 131-278.

882 94. Penzo E, Gandolfi G, Bargelloni L, Colombo L, Patarnello T., Messinian salinity crisis and the

883 origin of freshwater lifestyle in western Mediterranean Gobies. Mol Biol Evol. 1998;15: 1472-

884 1480.

885 95. Huyse T, Van Houdt J, Volckaert FAM. Paleoclimatic history and vicariant speciation in the

886 ‘sand goby’ group (Gobiidae, Teleostei). Mol Phylogenet Evol. 2004;32: 324-336.

887 96. Geiger MF , Herder F, Monaghan MT, Almada V, Barbieri R, Bariche M, et al. Spatial

888 heterogeneity in the Mediterranean Biodiversity Hotspot affects barcoding accuracy of its

889 freshwater fishes. Mol Ecol Resour. 2014;14: 1210-1221.

890 97. Gavrilets S. Perspective: models of speciation: what have we learned in 40 years? Evolution.

891 2003;57: 2197-2215.

892 98. Coyne J A, Orr HA. Speciation. Sinauer, Sunderland; 2004

893 99. Huttunen S, Aspi J, Schlotterer C, Routtu J, Hoikkala. A. Variation in male courtship song traits

894 in Drosophila virilis: the effects of selection and drift on song divergence at the intraspecific

895 level. Behav Genet. 2008;38: 82-92.

896 100. Campbell P, Pasch B, Pino JL, Crino OL, Phillips M, Phelps SM. Geographic variation in the

897 songs of neotropical singing mice: testing the relative importance of drift and local adaptation.

898 Evolution. 2010;64: 1955-1972.

899 101. Lachlan RF, Servedio MR. Song learning accelerates allopatric speciation. Evolution.

900 2004;58: 2049-2063.

901 102. Nosil P, Flaxman SM. Conditions for mutation-order speciation. Proc R Soc B: Biol Sci.

902 2011;278: 399-407.

903 103. Lynch A. The population memetics of birdsong. In: Kroodsma DE, Miller EH, editors.

904 Ecology and Evolution of Acoustic Communication in Birds. Cornell University Press; 1996.

905 pp. 181–197.

Predano na recenziju u Plos One

906 104. Rendella L, Whitehead H. Culture in whales and dolphins. Behav Brain Sci. 2001;24: 309-

907 382.

908 105. Schluter D. Evidence for ecological speciation and its alternative. Science. 2009;323: 737-

909 741.

910 106. Martin MD, Mendelson TC. Signal divergence is correlated with genetic distance and not

911 environmental differences in darters (Percidae: Etheostoma). Evol Biol. 2012;39: 231-241.

912 107. Lindner K, Cerwenka AF, Brandner J, Gertzen S, Borcherding J, Geist J, et al. First evidence

913 for interspecific hybridization between invasive goby species Neogobius fluviatilis and

914 Neogobius melanostomus (Teleostei: Gobiidae: Benthophilinae). J Fish Biol. 2013;82: 2128-

915 2134.

916 108. Seddon N, Tobias JA. Character displacement from the receiver’s perspective: species and

917 mate recognition despite convergent signals in suboscine birds. Proc Royal Soc. B. 2010;277:

918 2475-2483.

919 109. Pfennig KS, Pfennig DW. Character displacement: ecological and reproductive responses

920 to a common evolutionary problem. Q Rev Biol. 2009;84: 253-276.

921

922 Supporting information 923 924 Fig S1. Scatterplot from principle component analysis (PCA) performed with individuals means

925 of the six acoustic variables. PC1 is loaded by temporal variables number of pulses and sound

926 duration, while PC by spectral peak frequency and frequency modulation.

927

928 Fig S2. Correlation between genetic distance and acoustic divergence in nine vocal Gobius lineage

929 gobiids. In A), correlation was achieved (Mantel test r = 0.48, Pt.t. = 0.005) by using genetic distance

930 obtained from p-distance method for nuclear markers (rag1 and rho), while in B) correlation

931 (Mantel test r = 0.45, Pt.t. = 0.01) was inferred from mitochondrial cytb and cox1 sequences while

Predano na recenziju u Plos One

932 the divergence was obtained using p-distance method. The scatterplot represents the

933 relationship between species genetic differentiation and their acoustic distance.

934

935 Table S1. Partitioning scheme for concatenated dataset assessed by PartitionFinder 2.

936

937 Table S2. Mean values and standard deviations of the total length and the six acoustic variables

938 for the ten gobioid species. For Perccottus glenii, only thump sounds were used for the acoustic

939 analysis.

940

941 Table S3. Percentage and cumulative percentage of variance explained by the first three axis of

942 principal component analysis (PCA), with the loadings for these axes (i.e. factor coordinates)

943 extracted from six acoustic variables. PC factor coordinates represent the correlations between

944 the respective individual mean value of sound variable and each PC factor.

945

946

947

948

949

950

951

952

953

954

Predano na recenziju u Plos One

955 Figures

956

957 Fig. 1.

958

959 Fig. 2.

Predano na recenziju u Plos One

960

961 Fig. 3.

962

963 Fig. 4.

Predano na recenziju u Plos One

964

965 Fig. 5.

966

967 Fig. 6.

Predano na recenziju u Plos One

Correlation between acoustic divergence and phylogenetic distance in vocal European gobiids (Gobiidae; Gobius lineage)

Sven Horvatić1, Stefano Malavasi2*, Jasna Vukić3, Radek Šanda4, Zoran Marčić1, Marko Ćaleta5, Massimo Lorenzoni6, Perica Mustafić1, Ivana Buj1, Lucija Raguž1, Lucija Ivić1, Francesco Cavraro2 & Davor Zanella1*

1 Department of Zoology, Faculty of Science, University of Zagreb, Zagreb, Croatia

2 Department Environmental Sciences, Informatics and Statistics, Ca` Foscari University of Venice, Venezia Mestre, Italy

3 Department of Ecology, Charles University, Prague, Czech Republic

4 National Museum, Department of Zoology, Prague, Czech Republic

5 Faculty of Teacher Education, University of Zagreb, Zagreb, Croatia

6 Department of Chemistry, Biology and Biotechnologies, University of Perugia, Perugia, Italy

* Corresponding author

E-mail: [email protected] (DZ)

SUPPLEMENTARY INFORMATION

Table S1. Partitioning scheme for concatenated dataset assessed by PartitionFinder 2.

Substitution model Gene and codon position BI/RAxML HKY+I / GTR+G cox1-3, rho-1, cytb-1 TRN+G / GTR+G cytb-2, cox1-1 K80+G / GTR+G cytb-3 TRNEF+G / GTR+G cox1-2 K81UF+I / GTR+G rag1-1, rho-3 K80+G / GTR+G rag1-2 TVM / GTR+G rho-2, rag1-3

Predano na recenziju u Plos One

Table S2. Mean values and standard deviations of the total length and the six acoustic variables for the ten gobiod species. For Perccottus glenii, only thump sounds were used for the acoustic analysis.

Species N TL (mm) SR (s/min) DUR (ms) NP PRR (Hz) PF (Hz) FM (Hz) P. bonelli 5 75.6±7.9 18.2±6.6 679.2±145.7 57.6±5.6 90.6±7.9 134.2±11.4 -5.3±2.7 P. nigricans 4 94.7±10.2 24.0±3.7 260.6±66.5 19.9±7.4 76.1±13.2 89.4±15.0 1.8±2.8 G. paganellus 15 117.2±23.0 16.3±7.6 340.5±70.5 29.9±3.0 90.0±10.4 96.8±12.0 20.5±8.8 G. cobitis 4 147.5±37.4 3.3±1.2 330.3±97.3 16.5±7.8 48.6±8.8 86.4±12.1 -2.5±7.3 G. niger 5 121.4±12.4 8.4±4.3 368.2±32.6 16.4±2.5 44.7±5.9 109.4±11.2 -19.1±5.8 Z. ophiocephalus 8 175.7±30.8 6.3±2.6 253.9±66.2 9.2±1.8 36.9±3.1 215.8±11.0 -13.6±3.3 N. fluviatilis 7 130.6±10.8 4.4±5.5 170.0±39.5 12.4±2.5 73.7±6.9 78.0±5.8 8.8±4.7 N. melanostomus 7 145.5±12.5 13.4±11.7 128.2±35.6 10.0±1.6 83.5±17.9 83.3±18.8 3.8±6.9 P. kessleri 9 157.6±13.2 2.5±0.9 457.9±68.3 44.9±7.1 99.9±6.6 104.9±11.6 -9.1±10.6 P. glenii 6 106.1±5.4 9.2±3.9 95.4±10.4 8.7±1.2 92.4±7.8 97.9±10.0 / N, number of investigated individuals per species; TL, total length of the body (in millimeters). Abbreviations of the six acoustic variables: SR = sound rate, DUR = sound duration, NP = number of pulses, PRR = pulse repetition rate, PF = peak frequency, FM = frequency modulation.

Table S3. Percentage and cumulative percentage of variance explained by the first three axis of principal component analysis (PCA), with the loadings for these axes (i.e. factor coordinates) extracted from six acoustic variables. PC factor coordinates represent the correlations between the respective individual mean value of sound variable and each PC factor.

Variable PC1 PC2 PC3 Percentage 35.60 27.64 17.85 Cumulative percentage 35.60 63.25 81.10 SR (s/min) 0.31 -0.37 -0.75 DUR (ms) 0.73 0.56 -0.14 NP 0.96 0.24 0.05 PRR (Hz) 0.72 -0.42 0.30 PF (Hz) -0.22 0.69 -0.50 FM (Hz) 0.10 -0.71 -0.38

Predano na recenziju u Plos One

Fig S1. Scatterplot from principle component analysis (PCA) performed with individuals means of the six acoustic variables. PC1 is loaded by temporal variables number of pulses and sound duration, while PC by spectral peak frequency and frequency modulation.

Fig. S2. Correlation between genetic distance and acoustic divergence in nine vocal Gobius lineage gobiids. In A), correlation was achieved (Mantel test r = 0.48, Pt.t. = 0.005) by using genetic distance obtained from p-distance method for nuclear markers (rag1 and rho), while in B) correlation (Mantel test r = 0.45, Pt.t. = 0.01) was inferred from mitochondrial cytb and cox1 sequences while the divergence was obtained using p-distance method. The scatterplot represents the relationship between species genetic differentiation and their acoustic distance. DISKUSIJA

3. DISKUSIJA

XII DISKUSIJA DISKUSIJA

3.1. Glavoči linije Gobius – modelni organizmi u bioakustičkim istraživanjima

Svi su ciljevi obuhvaćeni ovom disertacijom izneseni u publikacijama I, II, III i IV koje su pokušale ispitati znanstvene hipoteze postavljene u okviru ove disertacije. Kao što je već navedeno, glavoči su jedna od najbrojnijih skupina vokalnih riba, s 24 dosad ispitane vrste u bioakustičkim eksperimentima, što čini oko 1,2% njihove ukupne brojnosti (Zeyl, 2016; Fricke i sur., 2020). Također, uz nekoliko drugih porodica (Pomacentridae, Batrachoidiidae, Gadidiae, Serrasalmidae), glavoči danas čine jednu od bioakustički najistraživanijih skupina riba koje se koriste u mnogim multidisciplinarnim bioakustičkim studijama kao što su akustička filogenija, fiziologija glasanja, vokalna etologija, utjecaj buke na akustičku komunikaciju i dr. (Amorim i Neves, 2007; Malavasi i sur., 2008; Parmentier i sur., 2013; Amorim i sur., 2018). Bez obzira na to, većina se znanstvene pozornosti i dalje pridaje tradicionalnim morfološko-ekološkim ispitivanjima, modernim molekularnim analizama radi rješavanja taksonomskih i sistematskih nepoznanica te dugoročnim projektima monitoringa (praćenja) glavoča određenog područja. Zbog svojih životnih navika i ekoloških karakteristika kao što su izražena teritorijalnost, bentički način života, speleofilni tip reprodukcije i dr., glavoči linije Gobius pokazali su se kao idealni laboratorijski (modelni) organizmi za eksperimente pasivne akustike ili in situ istraživanja. Za razliku od glavoča linije Pomatoschistus europskog područja (rodovi Knipowitschia, Pomatoschistus, Orsinigobius, Ninnigobius, Economidichthys), glavoči linije Gobius tjelesno su veći od navedenih skupina (do 27 cm ukupne dužine; Jardas, 1996), što ih čini iznimno pogodnim ribama za manipuliranje ili rukovanje tijekom laboratorijskih eksperimenata. Naime, pokazalo se kako se u eksperimentima pasivne akustike (publikacije I, II i IV) glavoče linije Gobius može, u laboratorijskim uvjetima, izazvati ili motivirati na produkciju zvukova jednostavnim, ali efektivnim etološkim „testom uljeza“ koji je postao standardni eksperimentalni postupak u bioakustičkim istraživanjima. Ukratko, u „testu uljeza“ određena jedinka domaćin (većinom mužjak) izolira se u posebnom akvariju na nekoliko sati/dana tijekom kojih razvija teritorijalno ponašanje te pri ubacivanju druge jedinke (istog ili različitog spola), počinje stvarati agresivne ili reproduktivne zvukove (publikacije I, II, III i IV i Sebastianutto i sur., 2008; Malavasi i sur., 2008, 2009). Dosad, samo kod jedne vrste glavoča, Economidichthys pygmaeus, zvukovi nisu zabilježeni u eksperimentima pasivne akustike tijekom reproduktivne sezone, tj. tijekom udvaranja ili parenja (Gkenas i sur., 2010). Rezultati navedenog etološko- akustičkog istraživanja navode kako je taj potencijalni sekundarni gubitak produkcije zvukova kod E. pygmaeus, inače jedne od najmanjih vrsta kralješnjaka na svijetu, najvjerojatnije

38 DISKUSIJA povezan s gubitkom kopanja ili gradnje gnijezda, što je kod većine vrsta uvijek popraćeno stvaranjem akustičkih signala (Gkenas i sur., 2010; Lugli, 2015). Ribe su u eksperimentima bioakustike vrlo popularni i u posljednjih dvadesetak godina često korišteni modelni organizmi, ponajprije zbog relativno jednostavne građe središnjih (unutarnje uho) i perifernih struktura za primanje i produkciju zvukova, relativno pojednostavljene strukture svojih akustičkih signala, manjeg vokalnog repertoara u odnosu na druge kralješnjake (uz nekoliko iznimki među ribama) te zbog snažne stereotipnosti i instinktivne prirode svojih zvukova (Malavasi i sur., 2008; Amorim i sur., 2011; Ladich, 2014). Uzimajući u obzir rezultate istraživanja koji pokazuju kako ribe (uključujući i glavoče) ne mogu naučiti zvukove ili ih oponašati onako kako to rade sisavci ili ptice, već su oni kod riba urođeni (Johnston i Buchanan, 2007; Longrie i sur., 2008), još jednom dolazimo do odgovora na pitanje zašto su upravo glavoči linije Gobius jedna od najčešće korištenih skupina riba u eksperimentima bioakustike. Uz to, velika brojnost validnih i opisanih vrsta iz porodice Gobiidae (oko 1.250 vrsta), kao i onih koje tek čekaju svoju taksonomsku sistematsku afilijaciju (oko 500 vrsta) (Fricke i sur., 2020), ostavlja mnogo prostora za nadolazeća bioakustička istraživanja srodnih vrsta glavoča linije Gobius šireg mediteranskog područja.

Krenuvši od prvog cilja, publikacije I i II blisko su povezane s proširivanjem općeg znanja o vokalnom repertoaru pontokaspijskih glavoča linije Gobius (Neogobius fluviatilis, N. melanostomus i Ponticola kessleri). Do trenutka objave navedenih publikacija, vokalni repertoar s opisanim i kvantificiranim akustičkim svojstvima, kao i etološki kontekst produkcije zvukova, bio je istražen kod samo dvije vrste pontokaspijskih glavoča, mramorastog glavoča Proterorhinus marmoratus (Pallas, 1814) i glavočića okrugljaka N. melanostomus (Ladich i Kratochvil, 1989; Rollo i sur., 2007). Mramorasti glavoč je slatkovodna i morska vrsta glavoča koja se donedavno smatrala široko rasprostranjenom u euroazijskim vodotocima, uključujući i rijeku Dunav (Miller, 2004), ali su molekularne analize pokazale kako se unutar Proterorhinus kompleksa nalazi nekoliko kriptičnih vrsta, s naglaskom na tome kako je slatkovodna forma najvjerojatnije odvojena vrsta, P. semilunaris (Heckel, 1837) (Neilson i Stepien, 2009b). Budući da su Ladich i Kratochvil (1989) opisali tonalne zvukove slatkovodne forme mramorastog glavoča (lokacija: rijeka Dunav kod Beča), najvjerojatnije se radi o vrsti P. semilunaris, kod koje je bioakustičko istraživanje pokazalo kako se tijekom agresivnih interakcija glasaju oba spola, dok se za vrijeme parenja glasaju samo mužjaci, stvarajući kratke (240 do 260 ms) tonalne zvukove frekvencije oko 70 – 130 Hz. Opisani zvukovi P. semilunaris po svojoj strukturi i izgledu spektrograma djelomično

39 DISKUSIJA nalikuju signalima dviju pontokaspijskih vrsta, N. fluviatilis i N. melanostomus, s kvantitativnim razlikama u vrijednostima akustičkih parametara (publikacija I i II). Također, vokalni repertoar P. kessleri, sa svojim akustičkim parametrima (publikacija II), omogućuje razlikovanje vrsta i skupina potporodice Benthophilinae s obzirom na akustičke signale. Neilson i Stepien (2009a) podijelili su, na temelju molekularnih analiza, grupu Benthophilinae na tri skupine (Neogobiini, Ponticolini, Bentophilini), a publikacije I i II grupiraju zabilježene kratke tonalne zvukove N. fluviatilis, N. melanostomus, koji se intenzivno brzo ponavljaju tijekom kraćeg razdoblja, u jednu grupu, dok dugi, frekvencijski modulirani tonalni zvukovi P. kessleri čine drugu skupinu zvukova (publikacija II). Umjereno dugi tonalni zvukovi P. semilunaris čine treću grupu, koja se nalazi između dvije već navedene skupine, što omogućuje razlikovanje glavoča skupine Neogobiini (N. fluviatilis, N. melanostomus; kratki i intenzivno ponovljeni tonalni zvukovi) od glavoča skupine Ponticolini (P. kessleri i P. semilunaris; umjereno dugi i dugi frekvencijsko modulirani tonalni zvukovi) s obzirom na strukturu akustičkog signala i njegove parametre (publikacija II). Nadalje, rezultati publikacije I i II pokazali su kako i dvije srodne vrste skupine Neogobiini (N. fluviatilis i N. melanostomus) dijele sličnu strukturu i izgled akustičkih signala (spektrogram i spektar snage), što ih jasno odvaja od ostalih vokalnih glavoča linije Gobius, iako i između navedene dvije vrste postoje jasne akustičke razlike odgovorne za njihovo međusobno odvajanje. Istovremeno, publikacije I i II potvrđuju kako grupiranje svih pontokaspijskih vrsta u jedinstven rod Neogobius (Berg, 1949) nije taksonomski ispravno te kako je rod Neogobius polifiletskog podrijetla, što su pokazali i rezultati molekularnih analiza (Neilson i Stepien, 2009a; Thacker i Roje, 2011; Medvedev, 2013).

Cilj publikacije I, uz opis akustičkih signala vrste N. fluviatilis, bio je istražiti kakvi su filogenetski odnosi između atlantsko-mediteranskih glavoča (rodovi Padogobius, Gobius i Zosterisessor) i navedene pontokaspijske vrste s obzirom na kvantitativne akustičke parametre. Prijašnja morfološka i molekularna istraživanja (Simonović, 1998; Thacker i Roje, 2011; Thacker, 2015) naglašavaju kako su pontokaspijski i atlantsko-mediteranski glavoči sestrinske skupine unutar linije Gobius, a rezultati publikacije I potvrdili su očekivani odnos budući da se N. fluviatilis grupirao na dijagramu s atlantsko-mediteranskim vrstama koje stvaraju isti tip zvuka i dijele slične akustičke parametre (P. nigricans, P. bonelli i G. paganellus). Ti rezultati odgovaraju na znanstveno pitanje „Kakvi su filogenetski odnosi pontokaspijskih i atlantsko-mediteranskih glavoča iz linije Gobius rekonstruirani na temelju kvalitativnih i kvantitativnih akustičkih parametara?“, naglašavajući kako između

40 DISKUSIJA pontokaspijske vrste i ostalih vokalnih vrsta linije Gobius postoji određena filogenetska povezanost s obzirom na akustička svojstva. Također, rezultati bioakustičke analize publikacije I navode kako su dva temporalna akustička parametra (trajanje – DUR i stopa ponavljanja pulseva – PRR, ali i maksimalna frekvencija zvuka – PF) najvjerojatnije zaslužna za zapaženo filogenetsko odvajanje istraživanih vokalnih glavoča. Takvi su rezultati poslužili kao osnova za daljnja bioakustička istraživanja radi proširivanja vokalnog repertoara neistraženih pontokaspijskih vrsta i ispitivanja filogenetskih odnosa pontokaspijskih s atlantsko-mediteranskim vrstama linije Gobius, s naglaskom na rod Padogobius koji se prijašnjim bioakustičkim istraživanjima (Lugli, 1995, 1996; Malavasi i sur., 2008), kao i publikacijom I, pokazao polifiletskim. Tu je hipotezu trebalo testirati molekularnim analizama, tj. upotrebom genetičkih biljega radi ispitivanja filogenetskih odnosa vokalnih vrsta linije Gobius, što će poslije u Diskusiji biti objašnjeno u sklopu publikacije IV. Još je jedan zanimljiv rezultat publikacije I, koji izravno odgovara na pitanja „Glasaju li se samo mužjaci tijekom intraspecijskih interakcija?“, a to je zapaženo i dokumentirano glasanje ženki N. fluviatilis tijekom agresivnih intraseksualnih (ženka – ženka) interakcija. Takva je pojava, tj. glasanje ženki tijekom agresivnih interakcija s drugim ženkama, zabilježena kod još jedne već spomenute pontokaspijske vrste, mramorastog glavoča P. semilunaris (Ladich i Kratochvil, 1989). Bioakustičko istraživanje publikacije I pokazalo je kako su tonalni zvukovi N. fluviatilis izrazito stereotipni i nepromijenjeni neovisno o etološkoj kategoriji produkcije zvukova (agresivna ili reproduktivna) te nije pronađena razlika po akustičkim parametrima i ostalim kvalitativnim svojstvima (izgled spektrograma i spektra snage) među spolovima. Navedeni rezultati publikacije I poklapaju se sa sličnim provedenim bioakustičkim istraživanjima (Myrberg i sur., 1965; Ladich, 1990, 2007; Lagardere i sur., 2005) koja navode kako se kod riba tijekom agresivnih interakcija oba spola glasaju, za razliku od reproduktivnih interakcija gdje je većinom mužjak taj koji se glasa, a njihovi se zvukovi rijetko razlikuju s obzirom na akustičke parametre budući da se zvukovi stvaraju tijekom obrane teritorija ili hranilišta (Ladich, 2015).

Ukratko, provedena bioakustička istraživanja u sklopu ove disertacije (publikacije I i II), zajedno s prijašnjim publikacijama (Ladich i Kratochvil, 1989; Lugli, 1995, 1996; Malavasi i sur., 2008, Malavasi i sur., 2009; Parmentier i sur., 2013), potvrđuju kako su glavoči linije Gobius, a pogotovo pontokaspijske vrste, izrazito dobri modelni organizmi za multidisciplinarna laboratorijska ispitivanja u bioakustičke svrhe. Oni nude velik znanstveni potencijal za buduće studije koje bi mogle odgovoriti na ključna pitanja iz područja

41 DISKUSIJA bioakustike riba („Zašto se ženke glavoča glasaju?“, „Koliko je zapravo vokalnih vrsta glavoča linije Gobius?“, „Kakav je utjecaj okolne buke na homeostazu glavoča?“).

3.2. Uvid u mehanizam za produkciju zvukova kod glavoča

Treći je cilj ove disertacije bio istražiti potencijalni mehanizam za produkciju zvukova kod rotana (P. glenii), pripadnika bazalne porodice Odontobutidae unutar podreda Gobioidei (filogenija prema Thacker, 2009). Taj je cilj blisko povezan s rezultatima publikacije III. Naime, kod većine riba glavnu ulogu u produkciji zvukova imaju posebni (sonički) mišići koji svojim uzastopnim kontrakcijama dovode do vibracije plivaćeg mjehura, koji se zbog toga smatra jednim od najvažnijih dijelova vokalnog sustava riba (Ladich i Fine, 2006; Fine i Parmentier, 2015). Zanimljivo je, doduše, kako kod glavoča možemo naći vokalne vrste koje imaju (Padogobius bonelli, Pomatoschistus minutus, Gobius cruentatus) ili nemaju plivaći mjehur (Padogobius nigricans, Neogobius melanostomus, N. fluviatilis), samim time odbacujući njegovu istaknutu ulogu u produkciji zvukova (Zeyl i sur., 2016). Nekoliko je potencijalnih mehanizama za produkciju zvukova predloženo kod glavoča, od soničkih mišića plivaćeg mjehura (Lugli i sur., 1995), hidrodinamičnog mehanizma temeljenog na rapidnim izbacivanjima vode iz otvora škržnog poklopca (Stadler, 2002) te kranio-pektoralnih mišića (Lugli i sur., 1996). Istražujući vokalni repertoar talijanskog glavoča Padogobius nigricans, Lugli (1996) navodi kako bi tonalni zvukovi te vrste mogli biti stvoreni serijom mišićnih kontrakcija posebnih mišića povezanih s oplećjem. Bioakustička analiza potvrdila je kako su tonalni zvukovi frekvencijsko-modulirani, na spektrogramu s izgledom sinusoidnog vala bez dodatnih harmonika iznad jednog glavnog (oko 100 Hz), što uz činjenicu kako P. nigricans nema plivaći mjehur potvrđuje kako zvukovi nisu stvoreni nekom drugom rezonantnom strukturom kao što je plivaći mjehur (Lugli i sur., 1996). Nadalje, rezultati navedenog istraživanja potvrđuju pretpostavku kako temperatura okoliša, tj. vodenog medija, ima znatan utjecaj na soničke mišiće koji najvjerojatnije sudjeluju u procesu stvaranja zvukova kod talijanskog P. nigricans (Lugli i sur., 1996). Kao što je navedeno u Uvodu, jedina su detaljna istraživanja mehanizma za produkciju zvukova glavoča proveli Parmentier i sur. (2013, 2017), u sklopu kojih su ispitani glavoči dviju različitih porodica, G. paganellus (Gobiidae) i Pomatoschistus pictus (Gobionellidae). Obje vrste, iako iz dviju različitih porodica, dijele identičan koštano-mišićni plan građe lubanje i oplećja te imaju sposobnost stvarati dva različitija tipa zvuka istim mehanizmom (G. paganellus – pulsatilni i tonalni zvuk; P. pictus – pulsatilni/bubnjajući i tupi zvuk). Autori pretpostavljaju kako se kod obje vrste sonički

42 DISKUSIJA mehanizam najvjerojatnije temelji na kontrakcijama kranio-pektoralnih mišića levator pectoralis (pars medialis i lateralis) koji povezuju lubanju s oplećjem (Parmentier i sur., 2013, 2017). Ti mišići imaju sve karakteristike brzokontrahirajućih mišića: vrpčaste miofibrile, razvijene tubule sakoplazmatskog retikuluma i razvijena područja s brojnim mitohondrijima u srži i periferiji (Parmentier i sur., 2013, 2017). Zanimljivo je napomenuti kako glavoči nemaju nikakve specifične mehaničke ili dodatne anatomske strukture, već upućuju na tipičan plan građe ostalih riba reda Perciformes (Lugli i sur., 1997; Malavasi i sur., 2008; Stadler, 2002), što pokazuje kako je mehanizam za produkciju zvukova mogao evoluirati od već postojećih struktura kroz proces egzaptacije (= iskorištavanje određene postojeće strukture za jednu funkciju kada je ona zapravo evoluirala za drugu, npr. perje ptica; Gould i Vrba, 1982). Uzimajući u obzir navedeni proces, autori pretpostavljaju kako je mehanizam za produkciju zvukova nastao od lokomotornih pokreta vezanih uz oplećje kod glavoča porodica Gobiidae i Gobionellidae (Parmentier i sur., 2013, 2017).

Jedino je istraživanje vokalnog repertoara i potencijalnog mehanizma za produkciju zvukova glavoča u širem smislu (Gobioidei) proveo Takemura (1983) koji je istražio produkciju zvukova i vokalne anatomske strukture uključene u proces akustičke komunikacije kod Odontobutis obscura, pripadnika porodice spavača (Odontobutidae). Rezultati istraživanja pokazali su kako mužjaci O. obscura stvaraju pulastilne zvukove u laboratorijskim uvjetima tijekom sezone reprodukcije (svibanj i lipanj), pri čemu se zvukovi sastoje od oko 8 pulseva trajanja 9 ms i maksimalne frekvencije ispod 1 kHz, s akustičkim parametrima (amplituda, raspon frekvencije i maksimalna frekvencija) znatno uvjetovanim veličinom ribe (veća riba – manja frekvencija i viša amplituda; Takemura, 1983). Anatomsko istraživanja O. obscura pokazalo je kako je stijenka plivaćeg mjehura izrazito tanka, bez izraženih mišića koji bi se vezali direktno na njega. Plivaći mjehur kod O. obscura nalazi se na dorzalnoj strani trbušne šupljine i ima izgled slova Y, s dva lateralna anteriorna proširenja oblika vilice koja završavaju ispod četvrtog kralješka. Dva posebna mišića, koji se anteriorno vežu na gornje ždrijelne zube, završavaju na 2. ̶ 4. kralješku i pokrivaju dva lateralna anteriorna proširenja plivaćeg mjehura (Takemura, 1983). Dodirivanjem ždrijelnog dijela manipulirane jedinke prstima, tijekom produkcije zvukova u laboratorijskim uvjetima, osjetile bi se slabe vibracije. Također, elektrostimulacija navedenih mišića tijekom eksperimentalne manipulacije dovela bi do produkcije zvukova ribe, s naglaskom na tome kako bi zvukovi postali izrazito tihi bušenjem ili uništavanjem plivaćeg mjehura, ali ne bi u potpunosti nestali. Iz svega navedenog, autor zaključuje kako se mehanizam za produkciju zvukova kod O.

43 DISKUSIJA obscura temelji najvjerojatnije na ždrijelnom mehanizmu, tj. nastanku zvukova struganjem gornjih i donjih ždrijelnih zubi, a njihova se amplifikacija događa zahvaljujući plivaćem mjehuru (Takemura, 1983).

Publikacija III istražila je etološki kontekst akustičke komunikacije i anatomske strukture povezane s potencijalnim mehanizmom za produkciju zvukova kod rotana Perccottus glenii. Znanstvena je hipoteza publikacije III glasila „Mehanizam za produkciju zvukova bazira se kod rotana Perccottus glenii (Odontobutidae) na kranio-pektoralnim mišićima, a pulsatilni zvukovi predstavljaju ancestralni/bazalni tip akustičkog signala kod glavoča u širem smislu“ s obzirom na hipotezu koju su predložili Malavasi i sur. (2008). Rezultati publikacije III blisko su povezani s prvim ciljem ove disertacije, tj. pokazali su kako mužjaci rotana, u laboratorijskim uvjetima, stvaraju dva tipa akustički različitih zvukova (tonalni i tupi) tijekom reproduktivne sezone, za vrijeme udvaranja (ženka izvan nastambe) ili tijekom interseksualnih interakcija u nastambi bez parenja. Zanimljivo je kako su tupi zvukovi (eng. thumps; < 200 ms i < 500 Hz, nemodulirani u frekvenciji) dokumentirani prvi put publikacijom III kod nekog pripadnika porodice Odontobutidae tijekom udvaranja ženki i njezina boravka u nastambi, dok su tonalni zvukovi (< 100 ms, 120 Hz, modulirani u frekvenciji) producirani samo tijekom udvaranja. Takav rezultat pokazuje kako oba zvuka potencijalno imaju različitu ulogu tijekom reproduktivne sezone, najvjerojatnije zbog svojih akustičkih parametara koji omogućuju frekvencijsko-moduliranim tonalnim zvukovima pokrivanje šireg spektra frekvencija u sluhu primatelja i širu transmisiju kroz okoliš, dok tupi zvukovi služe tijekom bliskih interakcija kada je ženka u nastambi. Slično su objašnjenje predložili Zeyl i sur. (2016), koji navode kako pulsatilni i tonalni zvukovi pokrivaju različite dijelove okolišnog spektra frekvencija budući da je kod sisavaca koji komuniciraju zvukovima na daljinu (npr. šišmiši i kitovi) modulacija frekvencije tonalnih zvukova česta akustička karakteristika (Wiley i Richards, 1982; Lugli i Fine, 2007). Treći je cilj publikacije III ispunjen etološkom analizom ponašanja vokalnih jedinki P. glenii, koja je potvrdila kako mužjaci produciraju tonalne zvukove pri bliskom kontaktu sa ženkama na ulazu u nastambu, a tupe zvukove tijekom interakcija izvan i unutar nastambe, s naglaskom na tome kako je većina (62%) etoloških kategorija popraćena tonalnim zvukovima, dok je kod tupih taj postotak znatno manji (23%). Anatomska analiza publikacije III, posebno blisko povezana s trećim ciljem disertacije, potvrdila je kako mužjaci P. glenii imaju koštano-mišićne dijelove lubanje i oplećja slične po svom rasporedu i strukturi onima kod ostalih glavoča, s manjim razlikama specifičnim za porodicu Odontobutidae (Winterbottom, 1974; Adriaens i sur.,

44 DISKUSIJA

1993; Li i sur., 2018). Naime, prijašnja su anatomska ispitivanja soničkog mehanizma glavoča (Parmentier i sur., 2013, 2017) istražila mišiće oplećja povezane s lubanjom (musculus levator pectoralis), čijim se kontrakcijama pokreće oplećje u ventro-anteriornom smjeru pri stvaranju zvukova tijekom eksperimentalnih manipulacija. Zajedno s opaženom odsutnošću pokreta bukalne i usne šupljine tijekom produkcije oba tipa zvuka kod P. glenii, rezultati publikacije III naglašavaju kako se mehanizam za stvaranje zvukova kod rotana najvjerojatnije temelji na kontrakcijama kranio-pektoralnih mišića levator pectoralis, koji se kod rotana dijele na tri dijela (pars lateralis superficialis, pars lateralis profundus i pars medialis), a na oplećju, tj. grlenjači, levator pectoralis pars medialis veže se dorzalno u odnosu na pars lateralis. Oba tipa zvukova (tonalni i tupi) kod P. glenii najvjerojatnije nastaju djelovanjem istog mehanizma koji se temelji na kontrakcijama kranio-pektoralnih levator pectoralis mišića, ali kako se točno zvukovi šire u okolišu te kako se amplificiraju, i dalje ostaje nerazjašnjeno. Mnoga istraživanja naglašavaju kako su sonički mišići poikilotermnih životinja (uključujući i ribe) pod znatnim utjecajem temperature vodenog medija (Feher i sur., 1998; Amorim i sur., 2015; Zeyl i sur., 2016), naglašavajući kako se akustički parametri znatno mijenjaju s povišenjem temperature vode zbog promjena u ritmu mišićnih kontrakcija. U slučaju P. glenii, oba su tipa zvukova, sa svojim akustičkim parametrima, bila pod znatnim utjecajem temperature vode (viša temperatura; viša frekvencija, viša stopa ponavljanja pulseva, veći broj pulseva, itd.), što potvrđuje kako upravo mišićne kontrakcije dovode do stvaranja zvukova i modulacije pojedinih parametara (publikacija III). Zanimljiv su koštani dio oplećja mnogih glavoča, uz navedene koštano-mišićne strukture lubanje, četiri velike radijalne kosti (ossa radiales) jače ili slabije povezane hrskavicom s kosti grljenjačom, koje su veće u odnosu na iste strukture drugih zrakoperki (Adriaens i sur., 1993). Naime, smatra se kako radijalne kosti, čineći jedinstvenu plohu nalik opni bubnja, imaju indirektnu funkciju u stvaranju zvukova kod G. paganellus i P. pictus budući da svojim vibracijama potencijalno amplificiraju zvukove stvorene kontrakcijama mišića levator pectoralis (Parmentier i sur., 2013, 2017). Koristeći analizu X-zraka u kombinaciji s mikrotomografijom radi istraživanja koštano-mišićnih svojstava oplećja spavača (Odontobutidae), Li i sur. (2018) navode kako se oplećje P. glenii sastoji od tipičnih dijelova za porodicu Odontobutidae, tj. imaju veliku, okoštanu lopaticu (os scapula) koja onemogućuje direktan kontakt proksimalnih radijalnih kostiju s grlenjačom (os cleithrum). Ako se budućim detaljnim multidisciplinarnim analizama (superbrze videokamere, elektromiografija, mikroskopija, itd.) potvrdi kako P. glenii dijeli morfološko- anatomske osobine i kinetičke kretnje lubanje i oplećja (klimanje ili lateralne undulacije) pri produkciji zvukova s istraženim vokalnim vrstama G. paganellus i P. pictus, onda će se moći

45 DISKUSIJA jasnije objasniti je li mehanizam za produkciju zvukova kod glavoča u širem smislu (Gobioidei) evolucijski konzerviran i smatra li se pleziomorfnim svojstvom za navedenu skupinu riba. Zasad, publikacija III potvrđuje kako glavoči (Gobiidae i Gobionellidae) dijele sa spavačima (Odontobutidae) sličan raspored i organizaciju kranio-pektoralnih mišića levator pectoralis, predloženih kao glavni dio mehanizma za produkciju zvukova glavoča u širem smislu (Parmentier i sur., 2013, 2017).

3.3. Akustička divergencija glavoča linije Gobius – povezanost s genetskom diferencijacijom?

Drugi i četvrti cilj ove disertacije blisko su povezani s ispitivanjem filogenetskih interspecijskih odnosa vokalnih glavoča linije Gobius, dok publikacije I i IV izravno odgovaraju na znanstvene hipoteze „Filogenetska interspecijska analiza pontokaspijskih i atlantsko-mediteranskih glavoča iz linije Gobius, temeljena na kvalitativnim i kvantitativnim akustičkim parametrima, upućuje na određen stupanj međusobne srodnosti između vokalnih vrsta. Također, pretpostavlja se kako se vokalni repertoar vokalnih glavoča linije Gobius sastoji od malog broja različitih tipova zvukova“ te „Interspecijski filogenetski odnosi devet vokalnih vrsta glavoča linije Gobius (rodovi Neogobius, Ponticola, Zosterisessor, Gobius, Padogobius), konstruirani na temelju šest kvantitativnih akustičkih parametara, preklapaju se s rezultatima molekularnih analiza, tj. zvukovi prenose filogenetski signal odgovoran za razlikovanje srodnih vrsta“. Generalno gledajući, filogenetski odnosi glavoča predstavljaju još uvijek neriješeni dio taksonomske zagonetke, vjerojatno zbog činjenice kako mnogi autori u svojim molekularnim istraživanjima koriste različite genetičke (jezgrene i mitohondrijske) biljege, raznovrsne metode filogenetske rekonstrukcije te brojne, ponekad nesrodne vrste, što dodatno otežava ionako kompliciranu situaciju sa sistematikom pojedinih vrsta ili linija. Cilj je publikacije I, a pogotovo publikacije IV, bio objediniti sve dotad postojeće podatke o bioakustici vokalnih glavoča linije Gobius te u početnom dijelu ove disertacije (publikacija I) istražiti postoji li bilo kakva akustička srodnost među ispitanim vrstama. Pozitivan je rezultat poslije poslužio kao osnova za dublja komparativna istraživanja (publikacija IV).

U publikaciji I, čiji je cilj bio istražiti akustičku srodnost pontokaspijske vrste N. fluviatilis s dotad dokumentiranim vokalnim atlantsko-mediteranskim glavočima (Lugli i sur., 1995, 1996, 1997; Malavasi i sur., 2008), pokazalo se kako bi zvukovi glavoča mogli poslužiti kao kvalitetan fenotipski karakter u ispitivanjima filogenetskih odnosa vokalnih vrsta linije Gobius. Publikacija I je, uz to što joj je cilj bio istražiti vokalni repertoar N. fluviatilis,

46 DISKUSIJA multivarijatnim statističkim pristupom ispitala i akustičku srodnost navedene vrste sa šest vrsta rodova Gobius, Zosterisessor i Padogobius. Rezultati akustičke analize potvrdili su kako su zvukovi istraživanih vrsta vokalnih glavoča izrazito specifični, a dodatna statistička analiza grupirala je vokalne jedinke (tj. njihove zvukove izražene s pet akustičkih parametara) u pripadajuće vrste s velikom razinom točnosti. Istovremeno, analiza je pokazala kako od nekoliko kvantitativnih akustičkih parametara samo temporalna (trajanje i stopa ponavljanja pulseva) i spektralna (maksimalna frekvencija) svojstva nisu korelirana s bilo kojim drugim parametrom, što je omogućilo kasnije korištenje u komparativnim analizama. Rezultati publikacije I pokazali su kako su se pojedine vrste smjestile na dijagramu u klastere s obzirom na sličnost i strukturu produciranog zvuka, a zanimljivo je kako su se vrste roda Padogobius (P. bonelli i P. nigricans) koje stvaraju tonalne zvukove grupirale s N. fluviatilis i G. paganellus, vokalnim glavočima koji produciraju također tonalne zvukove. Pogotovo je zanimljiv rezultat publikacije I kako su se dvije nesrodne vrste, P. nigricans i N. fluviatilis, grupirale zajedno s obzirom na akustičke parametre na dijagramu, što je pokazalo kako bi rod Padogobius mogao biti polifiletskog podrijetla s obzirom na to da je druga vrsta (P. bonelli) bila smještena na suprotnoj strani dijagrama u odnosu na P. nigricans i N. fluviatilis. Takav je rezultat, tj. polifiletsko podrijetlo roda Padogobius, predložen prijašnjim istraživanjima temeljenima na genetičkim biljezima (Penzo i sur., 1998; Huyse i sur., 2004). Stoga je publikacija I poslužila kao osnova za kasnija multidisciplinarna istraživanja temeljena na akustičko-molekularnim podacima koja su detaljnije istražila filogenetske odnose vokalnih glavoča linije Gobius i njihovu akustičku divergenciju (publikacija IV).

Zvukovi mnogih životinja, pogotovo ptica, žaba, sisavaca i kukaca, važno su komunikacijsko sredstvo za prijenos i izmjenu informacija, a koriste se tijekom agresivnih (borba za hranu, briga za potomstvo, signaliziranje o prisutnosti predatoru) i reproduktivnih (privlačenje partnera, udvaranje i parenje) interspecijskih ili intraspecijskih interakcija, zbog čega su kod nekih vokalnih skupina akustički signali povezani s njihovom evolucijom pa imaju izraženu taksonomsku vrijednost u razlikovanju vrsta, pa čak i populacija (Slabbekoorn, 2004; Velasquez i sur., 2013; Wilkins i sur., 2013). Kod pojedinih skupina, kao što su cikade i cvrčci, određene brzoevoluirajuće, ali fenotipski vrlo slične vrste mogu se odrediti samo na temelju međusobnih razlika u akustičkim signalima (Mendelson i Shaw, 2005; Marshall, 2008). Također, rezultati modernih komparativnih filogenetskih studija naglašavaju kako su akustički signali izrazito specifično fenotipsko svojstvo za vrstu ili rod te kako se morfološke/genetske razlike među vrstama mogu izraziti i razlikama u akustičkim signalima

47 DISKUSIJA

(Mendelson i Shaw, 2005; Marshall, 2008), što potvrđuje kako akustička divergencija među vrstama otkriva i upućuje na obrazac interspecijske diversifikacije unutar rodova (Seddon, 2008). Ti rezultati potvrđuju kako akustički signali imaju izraženu filogenetsku ulogu u divergenciji pojedinih vokalnih skupina, i u ranim (prezigotna izolacija, prije same reprodukcije) i u kasnim (post-zigotna izolacija) fazama specijacije (Wilkins i sur., 2013). Nadalje, pojedine publikacije naglašavaju kako su akustički signali izrazito bitno fenotipsko svojstvo kod simpatrijskih i srodnih vrsta (kao što su neke vrste glavoča rodova Pomatoschistus, Knipowitschia i Neogobius) koje ih stvaraju tijekom reproduktivnih interakcija budući da služe kao sredstvo filogenetske diskriminacije te na taj način izravno sudjeluju u procesu razdvajanja populacija i nastanka novih vrsta (Ritchie, 2007; Pedroso i sur., 2013; Wilkins i sur., 2013; Blom i sur., 2016; Zeyl i sur., 2016).

Cilj je publikacije IV istražiti filogenetsku srodnost vokalnih glavoča linije Gobius, rekonstruiranu na temelju DNK sekvenci četiriju genetičkih biljega (dva mitohondrijska: Citokrom b - cyt b i Citokrom oksidaza podjedinica I - COI i dva jezgrena: Rodopsin - Rho i Rekombinacijski aktivacijski gen 1 - RAG1) te dobivene rezultate usporediti s akustičkim podacima (tj. matricom interspecijske akustičke udaljenosti) kako bismo potvrdili postoji li poklapanje akustičke divergencije s genetičkom. Zanimljivo je kako se filogenetski odnosi mnogih životinja rekonstruirani na temelju akustičkih podataka često poklapaju s odnosima utvrđenima koristeći morfološke ili molekularne podatke, kao što je to slučaj kod kukaca, žaba, ptica i sisavaca (Oates i Trocco, 1983; Hoikkala i sur., 1994; Cocroft i Ryan, 1995; Desutter-Grandcolas, 1997; McCracken i Sheldon, 1997; Conner, 1999; Peters i Tonkin- Leyhausen, 1999; Robillard i Desutter-Grandcolas, 2004). Do danas niti jedna znanstvena publikacija nije kvantitativno istražila i empirijski usporedila, u smislu komparativne filogenetske analize, zvukove riba zrakoperki s molekularnim podacima, što je bio jedan od glavnih ciljeva publikacije IV. Uz to, publikacija IV ponudila je alternativnu hipotezu u odnosu na postojeću koju su predložili Malavasi i sur. (2008) povezanu s evolucijom akustičke divergencije glavoča linije Gobius (rodovi Gobius, Zosterisessor, Padogobius, Neogobius i Ponticola) s naglaskom na evolucijske sile (determinističke ili stohastičke) koje su potencijalno mogle dovesti do interspecijskog akustičkog razdvajanja istraživanih vrsta kao i opažene korelacije s molekularnim podacima. Rezultati publikacije IV potvrdili su kako se akustički signali istraživanih devet vokalnih glavoča linije Gobius međusobno razlikuju na temelju svih šest kvantitativnih akustičkih parametara, filogenetski su specifični za svaku vrstu te kao takvi omogućuju njihovo razlikovanje. Taj su rezultat potvrdile i multivarijatne

48 DISKUSIJA statističke analize koje su s velikim postotkom klasifikacije (oko 92%) grupirale zvukove pojedinih jedinki u odgovarajuće vrste te istaknule temporalne parametre zvuka (pogotovo trajanje – DUR i stopu ponavljanja pulseva – PRR, uz maksimalnu frekvenciju – PF) kao odgovorna svojstva za opaženo grupiranje/razdvajanje vrsta. Važno je napomenuti kako su publikacijom IV objedinjeni gotovo svi podaci povezani s vokalnim glavočima linije Gobius, dok su dvije dodatne vokalne vrste (Proterorhinus marmoratus/semilunaris i Gobius cruentatus) izostavljene zbog tehničkih razloga. Iz toga slijedi da je gotovo većina vokalnih glavoča linije Gobius bila uključena u istraživanje u sklopu publikacije IV, a inicijalna je bioakustička analiza pokazala kako bi pojedine vrste (P. nigricans s vrstama roda Neogobius) mogle biti srodnije nego što se to prije smatralo. Filogenetski odnosi prezentirani u obliku filograma (tj. filogenetskog stabla) i rekonstruirani na temelju četiriju genetičkih biljega koristeći metode maksimalne vjerodostojnosti i Bayesovski pristup (ML i BI), potvrdili su kako su pontokaspijske vrste roda Neogobius (N. fluviatilis i N. melanostomus) blisko povezane i srodne s talijanskim glavočem P. nigricans, s kojima još dijele i ekološko- morfološke karakteristike (nedostatak plivaćeg mjehura, raspored glavenih papila i kanala, stanište, itd.; Miller, 2003, 2004), dok je druga vrsta iz roda Padogobius (P. bonelli) bila smještena kao sestrinska vrsta svim ostalim istraživanim skupinama budući da se odvojila prva na filogenetskom (kombiniranom) stablu, odmah nakon vanjske grupe (P. glenii). Lugli i sur. (1996) naveli su, u sklopu bioakustičkog istraživanja zvukova talijanskog glavoča P. nigricans, kako bi dvije vrste roda Padogobius mogle biti filogenetski udaljene te kako se razlikuju po akustičkim signalima, što je poslije i potvrđeno molekularnim istraživanjima (Geiger i sur., 2014). Publikacija IV potvrdila je, uzimajući u obzir komparativne akustičke i molekularne rezultate, kako je rod Padogobius najvjerojatnije polifiletskog, dok su pontokaspijske vrste u cjelini gledano monofiletskog podrijetla. Za ostale vrste Gobius roda (G. paganellus, G. cobitis i G. niger), kao i za Z. ophiocephalus, filogenetski odnosi iz publikacije IV ostaju nerazriješeni na dubljoj taksonomskoj razini, što bi buduća molekularna istraživanja radi detaljnijeg rješavanja njihovih odnosa trebala riješiti uzimajući u obzir veći broj srodnih vrsta. Ipak, rezultati publikacije IV naglašavaju kako bi vrsta Z. ophiocephalus mogla biti srodna ostalim vrstama Gobius roda, pogotovo G. niger.

Mnoge multidisciplinarne studije navode kako akustički signali kod određenih skupina životinja prenose filogenetski signal zaslužan za interspecijsko razlikovanje vrsta, što potvrđuje opažena pozitivna kvantitativna korelacija između akustičke i genetičke divergencije (Päckert i sur., 2004; Percy i sur., 2006; Irwin i sur., 2008; Toews i Irwin, 2008;

49 DISKUSIJA

Thinh i sur., 2011; Velásquez i sur., 2013; Lee i sur., 2016). Ukratko, moderne komparativne studije navode kako se između filogenetski srodnijih vrsta može opaziti statistička ovisnost u vrijednostima njihovih filogenetskih osobina upravo zbog njihove bliže genetičke srodnosti (Felsenstein, 1985; Revell i sur., 2008), tj. tendencije srodnih vrsta da više nalikuju jedna drugoj nego vrstama nasumično izabranim s filogenetskog stabla, što je zapravo definicija filogenetskog signala (Blomberg i Garland, 2002; Münkemüller i sur, 2012). U sklopu publikacije IV komparativna akustičko-molekularna analiza potvrdila je kako postoji znatna korelacija između interspecijske genetske diversifikacije (uključujući matrice udaljenosti konstruirane za svaki pojedinačni gen ili sve gene zajedno) i akustičke divergencije, još jednom potvrđujući taksonomsku specifičnost akustičkih signala svake vokalne vrste glavoča linije Gobius iz ove disertacije. Taj rezultat pokazuje, prvi put dosad, kako se evolucija zvukova najvjerojatnije mijenjala (i dalje mijenja) u istom smjeru i jednakom stopom kao i genotip, bez obzira na tip i vrstu korištenih genetičkih biljega te kako stohastički procesi imaju dominantnu ulogu u oblikovanju opažene akustičke divergencije glavoča linije Gobius. Stoga rezultati publikacije IV izravno potvrđuju hipotezu „Interspecijski filogenetski odnosi devet vokalnih vrsta glavoča linije Gobius (rodovi Neogobius, Ponticola, Zosterisessor, Gobius, Padogobius), konstruirani na temelju šest kvantitativnih akustičkih parametara, preklapaju se s rezultatima molekularnih analiza, tj. zvukovi prenose filogenetski signal odgovoran za razlikovanje srodnih vrsta“. Također, iz opažene akustičko-molekularne korelacije slijedi kako zvukovi glavoča linije Gobius najvjerojatnije imaju filogenetski signal, kao što je to slučaj kod mnogih drugih vokalnih životinja (Freckleton i sur., 2002; Blomberg i sur., 2003; Escalona i sur., 2019) iako bi detaljna istraživanja trebala kvantitativno ispitati u kojoj je mjeri on prisutan u pojedinačnim temporalnim ili spektralnim parametrima zvuka.

Ova je disertacija istražila interspecijsku korelaciju akustičke divergencije s genetičkom diferencijacijom glavoča linije Gobius, što je zasad prvi pokušaj kvantificiranja filogenetske ovisnosti zvukova o genskoj divergenciji kod bilo koje skupine riba zrakoperki. Na temelju rezultata publikacija I, II, III i IV konstruirana je hipoteza o evoluciji zvukova istraživanih vokalnih glavoča linije Gobius, koja znatno proširuje hipotezu koju su predložili Malavasi i sur. (2008). Ti su autori predložili kako su najvjerojatnije upravo pulsatilni zvukovi predstavljali ancestralnu (zajedničku) karakteristiku svih glavoča, vjerojatno zbog činjenice kako je većina vrsta imala sposobnost stvoriti samo taj tip zvuka i kako je kod sestrinske skupine svih ostalih glavoča u širem smislu, Odontobutidae (sa predstavnikom O. obscura), zabilježen samo pulsatilni tip akustičkog signala (Takemura, 1983; Malavasi i sur.,

50 DISKUSIJA

2008). Dvije vrste iz sestrinske porodice svih ostalih glavoča, P. glenii i O. obscura (Odontobutidae), mogu producirati dva različita tipa zvukova (tonalni i tupi – pulsatilni). Iz toga slijedi kako su osnovni elementi za razvoj frekvencijsko-moduliranih i kompleksnih zvukova zabilježenih kod linije Gobius bili upravo kratki tonalni zvukovi (koji su prisutni kod pontokaspijskih vrsta Neogobius roda i talijanskog P. nigricans) te pulsatilni zvukovi (rodovi Gobius i Zosterisessor).

51 ZAKLJUČAK

4. ZAKLJUČAK

XIII ZAKLJUČAK ZAKLJUČAK

 Filogenetska interspecijska analiza pontokaspijskih i atlantsko-mediteranskih glavoča iz linije Gobius potvrdila je kako pojedine vokalne vrste iz ove disertacije dijele određen stupanj međusobne akustičke srodnosti (primjer vrste roda Neogobius s Padogobius nigricans i Gobius niger sa Zosterisessor ophiocephalus), dok se akustički repertoar glavoča iz linije Gobius sastoji od tri različita tipa zvukova.

 Tijekom intraspecijskih interakcija ne glasaju se samo mužjaci, već su i ženke vokalni spol (primjer N. fluviatilis).

 Sposobnost stvaranja zvukova zahvaljujući mehanizmu temeljenom na kontrakcijama kranio-pektoralnih mišića (mišići oplećja i lubanje) karakteristično je svojstvo za sve glavoče u širem smislu (Gobioidei). Također, za porodicu Odontobutidae (primjer P. glenii), zabilježena su dva tipa zvukova koji predstavljaju ancestralni vokalni repertoar glavoča u širem smislu (Gobioidei).

 Interspecijska akustička divergencija znatno je korelirana s genetičkom diferencijacijom glavoča linije Gobius te slijedi isti evolucijski obrazac bez obzira na tip genetičkog biljega (nDNA ili mtDNA), što pokazuje kako zvukovi prenose filogenetski signal zaslužan za razlikovanje srodnih vrsta linije Gobius.

52 LITERATURA

5. LITERATURA

53 LITERATURA LITERATURA

Adriaens D, Decleyre D, Verraes W (1993) Morphology of the pectoral girdle in Pomatoschistus lozanoi De Buen 1923 (Gobiidae) in relation to pectoral fin adduction. Belgian Journal of Zoology 123: 137-157. Agorreta A, San Mauro D, Schliewen U, Van Tassell JL, Kovačić M, Zardoya R, Rüber L (2013) Molecular phylogenetics of Gobioidei and phylogenetic placement of European gobies. Molecular Phylogenetic and Evolution 69: 619-633. Agorreta A, San Mauro D, Schliewen U, Van Tassell JL, Kovačić M, Zardoya R, Rüber L (2013) Molecular phylogenetics of Gobioidei and phylogenetic placement of European gobies. Molecular Phylogenetics and Evolution 69:619-633. Almada VC, Goncalves EJ, Oliveira RF, Almeida AJ, Santos RS, Wirtz P (2001) Patterns of diversity of the north-eastern Atlantic blenniid fish fauna (Pisces: Blenniidae). Global Ecology & Biogeography 10: 411-422. Amorim MCP (2006) Diversity in sound production in fish. U: Ladich F, Collin SP, Moller P, Kapoor BG (ed.) Communication in Fishes. Science Publishers Inc 71-105 Amorim MCP, Knight ME, Stratoudakis Y, Turner GF (2004) Differences in sounds made by courting males of three closely related Lake Malawi cichlid species. Journal of Fish Biology 65: 1358-1371. Amorim MCP, Neves ASM (2007) Acoustic signalling during courtship in the painted goby Pomatoschistus pictus. Journal of the Marine Biological Association of the United Kingdom 87: 1017-1023. Amorim MCP, Simões JM, Almada VC, Fonseca PJ (2011) Stereotypy and variation of the mating call in the Lusitaniantoadfish Halobatrachus didactylus. Behavioral Ecology and Sociobiology 65:707-716. Amorim MCP, Vasconcelos RO, Bolgan M, Pedroso SS, Fonseca PJ (2018) Acoustic communication in marine shallow waters: Testing the acoustic adaptive hypothesis in sand gobies. The Journal of Experimental Biology 221: jeb183681 Amorim MCP, Vasconcelos RO, Fonseca PJ (2015) Fish sounds and mate choice. U: Ladich F (ed.) Sound Communication in Fishes. Springer-Verlag 1-35 Au WWL, Hastings MC (2008) Principles of Marine Bioacoustics. Springer, New York Balls R (1948) Herring fishing with the echometer. Journal du Conseil / Conseil Permanent International pour l'Exploration de la Mer 15: 193-206. Banarescu P (1973) Origin and affinities of the freshwater fish fauna of Europe. Ichthyologia 5: 1-8. Barel CDN (1983) Towards a constructional morphology of cichlid fishes (Teleostei Perciformes). Netherlands Journal of Zoology 33: 357-424. Bass AH, Chagnaud BP (2012) Shared developmental and evolutionary origins for neural basis of vocal-acoustic and pectoral-gestural signaling. Proceedings of the National Academy of Sciences of the United States of America 109: 10677-10684.

54 LITERATURA

Beeman K (1998) Digital Signal Analysis Editing and Synthesis. U: Hopp SL, Owren MJ, Evans CS (ed.) Animal Acoustic Communication. Springer Berlin Heidelberg Bellwood DR, Wainwright PC (2002) The history and biogeography of fishes on coral reefs. U: Sale PF (ed.) Coral reef fishes: dynamics and diversity in a complex ecosystem. Academic Press San Diego Berg LS (1949) Freshwater Fishes of the USSR and Adjacent Countries. Natuurkundig tijdschrift voor Nederlandsch Indië 5: 317-352. Berra T (2001) Freshwater fish distribution. Academic Press San Diego California Betancur-R R, Broughton RE, Wiley EO, Carpenter K, Lopez JA, Li C Holcroft NI, Arcila D, Sanciangco M, Cureton JC, Zhang F, Buser T, Campbell MA, Ballesteros JA, Roa-Varon A, W SW, Borden C, Rowley T, Reneau TC, Hough DJ, Lu G, Grande T, Arratia G, Ortí G (2013) The tree of life and a new classification of bony fishes. PLOS Currents Tree of Life, Edition 1. doi: 10.1371/currents.tol.53ba26640df0ccaee75bb165c8c26288. Betancur-R R, Wiley EO, Arratia G, Acero A, Bailly N, Miya M, Lecointre G, Ortí G (2017) Phylogenetic classification of bony fishes. BMC Evolutionary Biology 17: 162. Blom EL, Mück I, Heubel K, Svensson O (2016) Acoustic and visual courtship traits in two sympatric marine Gobiidae species—Pomatoschistus microps and Pomatoschistus minutus. Environtal Biology of Fishes 99: 999-1007. Blomberg SP, Garland T (2002) Tempo and mode in evolution: phylogenetic inertia adaptation and comparative methods. Journal of Evolutionary Biology 15: 899-910. Boyle KS, Dewan AK, Tricas TC (2013) Fast drum strokes: novel and convergent features of sonic muscle ultrastructure innervation and motor neuron organization in the pyramid butterflyfish (Hemitaurichthys polylepis). Journal of Morphology 274: 377-394. Briggs JC (1974) Marine Zoogeography. McGraw-Hill New York. Charlebois PM, Marsden JE, Goettel RGR, Wolfe RK, Jude DJ, Rudnika S (1997) The Round Goby Neogobius melanostomus (Pallas) A Review of European and North American Literature. Illinois-Indiana Sea Grant Program and Illinois Natural History Survey Illinois. INHS Special Publication Christensen CB, Christensen-Dalsgaard J, Madsen PT (2015) Hearing of the African lungfish (Protopterus annectens) suggests underwater pressure detection and rudimentary aerial hearing in early tetrapods. The Journal of Experimental Biology 218: 381-387 Cocroft RB, Ryan MJ (1995) Patterns of advertisement call evolution in toads and chorus frogs. Animal Behaviour 49: 283-303. Cohen DM, Robins CR (1970) A new ophidioid fish (genus Lucifuga) from a limestone sink. New Providence Island Bahamas. Proceedings of the Biological Society of Washington 83: 133-144. Colleye O, Ovidio M, Salmon A, Parmentier E (2013) Contribution to the study of acoustic communication in two Belgian river bullheads (Cottus rhenanus and C perifretum) with further insight into the sound-producing mechanism. Frontiers in Zoology 10:71

55 LITERATURA

Colleye O, Parmentier E (2012) Overview of the diversity of sounds produced by the clownfishes (Pomacentridae): importance of acoustic signals in their peculiar way of life. PLoS ONE 7: 1-11 Conner WE (1999) ‘Un chant d’appel amoreux’: acoustic communication in moths. Journal of Experimental Biology 202: 1711-1723. Cowman PF (2014) Historical factors that have shaped the evolution of tropical reef fishes: a review of phylogenies biogeography and remaining questions. Frontiers in Genetics 5: 394. De Jong K, Amorim MCP, Fonseca PJ, Fox CJ, Heubel KU (2018) Noise can affect acoustic communication and subsequent spawning success in fish. Environmental Pollution: 237:814- 823 Desutter-Grandcolas L (1997) A phylogenetic analysis of the evolution of the stridulatory apparatus in true crickets (Orthoptera Grylloidea). Caldistics 13: 101-108. Escalona DSM, Simões IP, Gonzalez-Voyer A, Castroviejo-Fisher S (2019) Neotropical frogs and mating songs: The evolution of advertisement calls in glassfrogs. Journal of Evolutionary Biology 32: 163-176. Fay RR (1988) Hearing in Vertebrates: A Psychophysics Databook Winnetka. Hill-Fay Associates Feher J, Waybright T, Fine M (1998) Comparison of sarcoplasmic reticulum capabilities in toadfish (Opsanus tau) sonic muscle and rat fast twitch muscle. Journal of Muscle Research and Cell Motilityil 19: 661-674. Felsenstein J (1985) Phylogenies and the comparative method. American Naturalist 125: 1-15. Fine M L, Parmentier E (2015) Mechanisms of fish sound production. U: Ladich F (ed.) Sound Communication in Fishes. Springer-Verlag Wien 77-126 Fish MP, Mowbray WH (1970) Sounds of Western North Atlantic fishes. Johns Hopkins Press Baltimore Freckleton RP, Harvey PH, Pagel M (2002) Phylogenetic analysis and comparative data: a test and review of evidence. American Naturalist 160: 712-726. Freyhof J (2011) Diversity and Distribution of Freshwater Gobies from the Mediterranean the Black and Caspian Seas. U: Patzner RA, Van Tassell JL, Kovačić M, Kapoor BG (ed.) The Biology Of Gobies. Science Publishers CRC Press 279-288 Fricke R, Eschmeyer WN, Van der Laan R (2020) Eschmeyer’s catalog of fishes: genera species references. Digital resource at http://researcharchive.calacademy.org/research/ichthyology/catalog/fishcatmain.asp. California Acadamey of Science Geiger MF, Herder F, Monaghan MT, Almada V, Barbieri R, Bariche M, Berrebi P, Bohlen J, Casal-Lopez M, Delmastro GB, Denys GP, Dettai A, Doadrio I, Kalogianni E, Kärst H, Kottelat M, Kovačić M, Laporte M, Lorenzoni M, Marčić Z, Özuluğ M, Perdices A, Perea S, Persat H, Porcelotti S, Puzzi C, Robalo J, Šanda R, Schneider M, Šlechtová V, Stoumboudi M, Walter S, Freyhof J (2014) Spatial heterogeneity in the Mediterranean Biodiversity

56 LITERATURA

Hotspot affects barcoding accuracy of its freshwater fishes. Molecular Ecology Resources 14: 1210-21. Gierl C, Reichenbacher B, Gaudant J, Erpenbeck D, Pharisant A (2013) An extraordinary gobioid fish fossil from southern France. PLoS ONE 8: e64117. Gkenas C, Malavasi S, Georgalas V, Leonardos ID, Torricelli P (2010) The reproductive behavior of Economidichthys pygmaeus : secondary loss of sound production within the sand goby group? Environmental Biology of Fishes 87: 299-307 Gould SJ, Vrba ES (1982) Exaptation-a missing term in the science of form. Paleobiology 8: 4-15. Hawkins AD (1993) Underwater sound and fish behaviour U: Pitcher TJ (ed.) Behaviour of Teleost Fishes. Chapman Hall London 129-169 Hawkins AD, Myrberg AAJ (1983) Hearing and sound communication underwater. U: Lewis B (ed.) Bioacoustics, A Comparative Approach. Academic Press 347-405 Henglmüller SM, Ladich F (1999) Development of agonistic behaviour and vocalization in croaking gourami. Journal of Fish Biology 54: 380-395 Hoese D F (1984) Gobioidei: relationships. U: Moser HG (ed.) Ontogeny and Systematics of Fishes. Special Publication of the American Society of Ichthyologists and Herpetologists, Allen Press Lawrence Kansas Hoese DF, Gill AC (1993) Phylogenetic relationships of eleotridid fishes (Perciformes: Gobioidei). Bulletin of Marine Science 52: 415-440. Hoikkala A, Kaneshiro KY, Hoy RR (1994) Courtship songs of the picture-winged Drosophila planitibia subgroup species. Animal Behaviour 47: 1363-1374 Holt DE, Johnston CE (2014) Evidence of the Lombard effect in fishes. Behavioral Ecology 25: 819-826 Horn MH (1972) The amount of space available for marine and freshwater fishes. Fishery Bulletin 70: 1295-1297. Hsü KJ, Montadert L, Bernoulli D, Cita MB, Erickson A, Garrison RE, Kidd RB, Melieres F, Muller C, Wright R (1977) History of the Mediterranean salinity crisis. Nature 267: 399-403. Huyse T, Van Houdt J, Volckaert AM (2004) Paleoclimatic history and vicariant speciation in the ‘sand goby’ group (Gobiidae Teleostei). Moleular Phylogenetics and Evolution 32: 324- 336. Irwin D, Thimgan M, Irwin J (2008) Call divergence is correlated with geographic and genetic distance in Greenish Warblers (Phylloscopus trochiloides): a strong role for stochasticity in signal evolution? Journal of Evolutionary Biology 21: 435-48. Isabella‐Valenzi L, Higgs DM (2016) Development of an acoustic trap for potential round goby (Neogobius melanostomus) management. Journal of Great Lakes Research 42: 904-909. Jardas I (1996) Jadranska Ihtiofauna. Školska knjiga Zagreb 331-359

57 LITERATURA

Johnson GD, Brothers EB (1993) Schindleria: a paedomorphic goby (Teleostei: Gobioidei). Bulletin of Marine Science 52: 441-471. Johnston CE, Buchanan HM (2007) Learned or innate production of acoustic signals in fishes: a test using a cyprinid. Environmental Biology of Fish 78: 183-187. Johnston CE, Johnson DL (2000) Sound production during the spawning season in cavity- nesting darters of the subgenus Catonotus (Percidae: Etheostoma). Copeia 2000: 475-481. Karplus I (1979) The tactile communication between Cryptocentrus steinitzi (Pisces Gobiidae) and Alpheus purpurilenticularis (Crustacea Alpheiidae). Zeitschrift für Tierpsychologie 49: 173-196. Kihslinger RL, Klimley AP (2002) Species identity and the temporal characteristics of fish acoustic signals. Journal of Comparative Psychology 116: 210-214. Kottelat M, Freyhof J (2007) Handbook of European Freshwater Fishes. Publications Kottelat Cornol Kovačić M, Patzner RA (2011) North-Eastern Atlantic and Mediterranean gobies. U: Patzner RA, Van Tassell JL, Kovačić M, Kapoor BG (ed.) The Biology Of Gobies. Science Publishers CRC Press 177-193 Kovačić M, Svensen R (2019) Northern extension of Lesueurigobius friesii (Malm, 1874) (Pisces: Gobiidae) distribution and the gobiid diversity decline along the Norwegian coast. Acta Adriatica 60: 147-156. Kovačić M (2020) Checklist of gobies (Teleostei: Gobiidae) of the Mediterranean Sea and a key for species identification. Zootaxa 4877: 75-101. Krijgsman W, Hilgen FJ, Raffi I, Sierro FJ, Wilson DS (1999) Chronology causes and progression of the Messinian salinity crisis. Nature 400: 652-655. Ladich F (1990) Vocalization during agonistic behaviour in Cottus gobio L (Cottidae): an acoustic threat display. Ethology 84: 193-201. Ladich F (2000) Acoustic communication and the evolution of hearing in fishes. Philosophical Transactions of the Royal Society B: Biological Sciences 355: 1285-1288 Ladich F (2001) Sound-generating and -detecting motor system in catfish: design of swimbladder muscles in doradids and pimelodids. The Anatomical Record 263: 297-306. Ladich F (2004) Sound production and acoustic communication U: von der Emde G, Mogdans J, Kapoor BG (ed.) The senses of fish: adaptations for the reception of natural stimuli. Kluwer Dordrecht 210-230 Ladich F (2007) Females whisper briefly during sex: context- and sex-specific differences in sounds made by croaking gouramis. Animal Behav 73: 379-387. Ladich F (2014) Fish bioacoustics. Current Opinion in Neurobiology 28 121-127 Ladich F (2015) Sound Communication in Fishes. Vienna: Springer Ladich F (2019) Ecology of sound communication infishes. Fish and Fisheries 20: 552 563

58 LITERATURA

Ladich F, Fine M L (2006) Sound-generating mechanisms in fishes: a unique diversity in vertebrates. U: Ladich F, Collin SP,Moller P, Kapoor BG (ed.) Communication in Fishes. Science Publishers 3-43 Ladich F, Fine ML (2006) Sound-generating mechanisms in fishes: a unique diversity in vertebrates. U: Collin SP, Moller P, Kapoor BG Ladich F (ed.) Communication in fishes. Science Publishers 3-43. Ladich F, Kratochvil H (1989) Sound production in the Marmoreal goby Proterorhinus marmoratus (Pallas) (Gobiidae: Teleostei). Zoologische Jahrbucher Abteilung fuer Allgemeine Zoologie und Physiologie der Tiere 93: 501-504. Ladich F, Popper AN (2004) Parallel evolution in fish hearing organs U: Manley G, Popper AN, Fay RR (ed.) Evolution of the Vertebrate auditory system. Springer New York 95-127 Ladich F, Schulz-Mirbach T (2016) Diversity in fish auditory systems: One of the riddles of sensory biology. Frontiers in Ecology and Evolution 4: 28 Ladich F, Winkler H (2017) Acoustic communication in terrestrial and aquatic vertebrates. The Journal of Experimental Biology 220: 2306-2317 Lagardere JP, Millot S, Parmentier E (2005) Aspects of sound communication in the pearlfish Carapus boraborensis and Carapus homei (Carapidae). The Journal of Experimental Biology 303: 1066-1074. Larson HK (1995) A review of the AustraHan endemic gobiid fish genus Chlamydogobius with description of five new species. The Beagle, Records of the Museums and Art Galleries of the Northern Territory 12: 19-51. Lee KH, Shaner PJ, Lin YP, Lin SM (2016) Geographic variation in advertisement calls of a Microhylid frog - testing the role of drift and ecology. Molecular Ecology 6: 3289-3298. Li H, He Y, Jiang J, Liu Z, Li C (2018) Molecular systematics and phylogenetic analysisof the Asian endemic freshwater sleepers (Gobiiformes: Odontobutidae). Molecular Phylogenetics and Evolution 121: 1-11. Liem KF (1993) Ecomorphology of the teleostean skull U: J Hankenj, Hall BK (ed.) The Skull: Functional and Evolutionary Mechanisms. University of Chicago Press 422-452. Limbaugh C (1961) Cleaning symbiosis. Scientific American 205: 42-49. Lindström K, Hellström M (1993) Male size and parental care in the sand goby Pomatoschistus minutus. Ethology Ecology & Evolution 5: 97-106. Linnaeus C (1758) Systema Naturae per regna tria naturae, secundum classes, ordines, genera, species, cum characteribus, differentiis, synonymis, locis vol. 1. 10th edn. Laurentius Salvius, Holmiae Lobel PS (1998) Possibile species specific courtship sounds by two sympatric cichlids fishes in Lake Malati Africa. Environmental Biology of Fishes 52: 443-452. Loesser KE, Rafi J, Fine ML (1997) Embryonic juvenile and adult development of the toadfish sonic muscle. The Anatomical Record 249: 469-477.

59 LITERATURA

Longrie N, Fine ML, Parmentier E (2008) Innate sound production in the cichlid Oreochromis niloticus. Journal of Zool 276: 266-275. Losey GS Jr (1987) Cleaning symbiosis. Symbiosis 4: 229-258. Luczkovich JJ, Mann DA, Rountree RA (2008) Passive acoustics as a tool in fisheries science. Transactions of the American Fisheries Society 137: 533-541 Lugli M (2010) Sounds of shallow water fi shes pitch within the quiet window of the habitat ambient noise. Journal of Comparative Physiology A 196: 439-196, Lugli M (2012) Acoustics of fish shelters: frequency response and gain properties Journal of the Acoustical Society of America 132: 3512-3524 Lugli M (2013) Sand pile above the nest amplifies the sound emitted by the male sand goby Environmental Biology of Fishes 96: 1003-1012 Lugli M (2014) Acoustics of fi sh shelters: background noise and signal-to-noise ratio. Journal of the Acoustical Society of America 136: 3382-3388 Lugli M (2015) Habitat acoustics and the low-frequency communication of shallow water fishes. U: Ladich F (ed.) Sound Communication in Fishes. Springer-Verlag Wien Lugli M, Fine ML (2007) Stream ambient noise spectrum and propagation of sounds in the goby Padogobius martensii: Sound pressure and particle velocity. Journal of the Acoustical Society of America 122: 2881-2892. Lugli M, Pavan G, Torricelli P, Bobbio L (1995) Spawning vocalizations in male freshwater gobies (Pisces Gobiidae). Environmental Biology of Fishes 43: 219-231. Lugli M, Torricelli P (1999) Prespawning sound production in Mediterranean sand-gobies. Journal of Fish Biology 54: 691-694 Lugli M, Torricelli P, Pavan G, Mainardi D (1997) Sound production during courtship and spawning among freshwater gobiids (Pisces Gobiidae). Marine and Freshwater Behaviour and Physiology 29: 109-126 Lugli M, Torricelli P, Pavan G, Miller PJ (1996) Breeding sounds of male Padogobius nigricans (Teleostei: Gobiidae) with suggestions for further evolutionary study of vocal behaviour in gobioid fishes. Journal of Fish Biology 49: 648-657. Lugli M, Yan HY, Fine ML (2003) Acoustic communication in two freshwater gobies: the relationship between ambient noise hearing thresholds and sound spectrum. Journal of Comparative Physiology A 189: 309-320 Malavasi S, Collatuzzo S, Torricelli P (2008) Interspecific variation of acoustic signals in Mediterranean gobies (Perciformes Gobiidae): comparative analysis and evolutionary outlook. Biological Journal of the Linnean Society 93: 763-778. Malavasi S, Torricelli P, Lugli M, Pranovi F, Mainardi D (2003) Male Courtship Sounds in a Teleost with Alternative Reproductive Tactics the Grass Goby Zosterisessor ophiocephalus. Environmental Biology of Fishes 66: 231-236.

60 LITERATURA

Malavasi S, Valerio C, Torricelli P (2009) Courtship sounds and associated behaviours in the Canestrini’s goby Pomatoschistus canestrinii. Journal of Fish Biology 75: 1883-1887. Mann DA (2006) Propagation of fi sh sounds U: Ladich F, Collin S,, Moller P Kapoor B (ed.) Communication in fishes Vol I. Science Publishers New Hampshire 107-120 Mann DA, Hawkins AD, Jech JM (2008) Active and passive acoustics to locate and study fish. U: Webb JF, Fay RR, Popper AN (ed.). Fish bioacoustics. Springer New York 279-309 Mann DA, Lobel PS (1995) Passive acoustic detection of sounds produced by the damselfish Dascyllus albisella (Pomacentridae). Bioacoustics 6: 199-213. Marshall DC, Slon K, Cooley JR, Hill KBR, Simon C (2008) Steady Plio-Pleistocene diversification and a 2-million-year sympatry threshold ina New Zealand cicada radiation. Molecular Phylogenetics and Evolution 48: 1054-1066. McCracken KG, Sheldon FH (1997) Avian vocalizations and phylogenetic signal. Proceedings of the National Academy of Sciences of the United States of America 94: 3833- 3836. Mckay SI, Miller PJ (1991) Isozyme criteria in the testing of phyletic relationships between species of Gobius and related eastern Atlantic-Mediterranean genera (Teleostei: Gobiidae). Journal of Fish Biology 39: 291-299. McKay SI, Miller PJ (1991) Isozyme criteria in the testing of phyletic relationships between species of Gobius and related eastern Atlantic-Mediteranean genera (Teleostei: Gobiidae). Journal of Fish Biology 39: 291-299 McKay SI, Miller PJ (1997) The affinities of European sand gobies (Teleostei: Gobiidae). Journal of Natural History 31: 1457-1482. Medvedev DA, Sorokin PA, Vasil’ev VP, Chernova NV, Vasil’eva ED (2013) Reconstruction of phylogenetic relations of Ponto-Caspian gobies (Gobiidae Perciformes) based on mitochondrial genome variation and some problems of their taxonomy. Journal of Ichthyology 53: 702-712. Mendelson TC, Shaw KL (2005) Sexual behaviour: rapid speciation in an arthropod. Nature 433: 375-376 Miller PJ (1973) The osteology and adaptive features of Rhyacichthys aspro (Teleostei: Gobioidei) and the classification of gobioid fishes. Journal of Zoology London 171:397-434. Miller PJ (1979) Adaptiveness and implications of small size in teleosts. Symposia of the Zoological Society of London 44: 263-306. Miller PJ (1986) Gobiidae. U: Whitehead PJP, Bauchot ML, Hureau JC, Nielsen J, Tortonese E (ed.). Fishes of the North-eastern Atlantic and the Mediterranean, 3. UNESCO Paris 1019- 1085. Miller PJ (1990) The endurance of endemism: the Mediterranean freshwater gobies and their prospects for survival. Journal of Fish Biology 37: 145-156. Miller PJ (1996) The functional ecology of small fish: some opportunities and consequences. Symposia of the Zoological Society of London 69: 175-199.

61 LITERATURA

Miller PJ (2003) Gobiidae. U: Miller PJ (ed.). The Freshwater Fishes of Europe Vol. 8/I. Aula Verlag Wiebelsheim Miller PJ (2004) Gobiidae 2. U: Miller PJ (ed.). The Freshwater Fishes of Europe Vol. 8/II. Aula Verlag Wiebelsheim Miller PJ, Serventi M, Soregaroli D, Torricelli P, Gandolfi G (1994) Isozyme genetics and the phylogeny of Italian freshwater gobies (Teleostei: Gobioidei). Journal of Fish Biology. 44:439-451. Münkemüller T, Lavergne S, Bzeznik B, Dray S, Jom-bart T, Schiffers K, Thuiller W (2012) How to measure and test phylogenetic signal. Methods in Ecology and Evolution 3: 743- 756. Myrberg AA, Kramer E, Heinecke P (1965) Sound production by cichlid fishes. Science 149: 555-558 Myrberg AA, Lugli M (2006) Reproductive behavior and acoustical interactions U: Ladich F, Collin SP, Moller P, Kapoor BG (ed.) Communication in fishes vol I. Science Publishers 149- 176 Near TJ Dornburg A Eytan RI Keck BP Smith WL Kuhn KL Moore JA Price SA Burbrink FT Friedman M Wainwright PC (2013) Phylogeny and tempo of diversification in the superradiation of spinyrayed fishes. Proceedings of the National Academy of Sciences USA 110: 12738-21743. Neilson ME, Stepien CA (2009a) Escape from the Ponto-Caspian: evolution and biogeography of an endemic goby species flock (Benthophilinae: Gobiidae: Teleostei). Molecular Phylogenetics and Evolution 52: 84-102 Neilson ME, Stepien CA (2009b) Evolution and phylogeography of the tubenose goby genus Proterorhinus (Gobiidae: Teleostei): evidence for new cryptic species. Biological Journal of the Linnean Society 96: 664-684. Nelson JS Grande TC Wilson MV (2006) Fishes of the World. 4th edn. John Wiley & Sons New Jersey Nelson JS Grande TC Wilson MV (2016) Fishes of the World. 5th edn. John Wiley & Sons New Jersey Nursall JR (1981) Behavior and habitat affecting the distribution of five species of sympatric mudskippers in Queensland. Bulletin of Marine Science 31: 730-735. Oates JF, Trocco TF (1983) Taxonomy and phylogeny of black and white colobus monkeys Inferences from analysis of loud call variation. Folia Primatologica 40: 83-113. Otovic P, Partan S (2009) Multimodal signaling in animals U: Squire LR (ed.) Encyclopedia of neuroscience London (UK). Academic Press 1095-1105 Päckert M, Martens J, Sun Y, Veith M (2004) The radiation of the Seicercus burkii complex and its congeners (Aves: Sylviidae): molecular genetics and bioacoustics. Organisms Diversity & Evolution 4: 341-364.

62 LITERATURA

Parmentier E, Di Iorio L, Picciulin M, Malavasi S, Lagardère J-P, Bertucci F (2017) Consistency of spatiotemporal sound features supports the use of passive acoustics for long-term monitoring. Animal Conservation 21: 211-220. Parmentier E, Fine M L (2016) Fish sound production: insights. U: R Suthers R, Fitch W, Fay R, Popper A (ed.) Vertebrate Sound Production and Acoustic Communication. Springer 19-49 Parmentier E, Kéver L, Boyle K, Corbisier Y, Sawelew L, Malavasi S (2013) Sound production mechanism in Gobius paganellus (Gobiidae). The Journal of Experimental Biology 216: 3189-3199. Parmentier E, Kéver L, Boyle K, Corbisier Y, Sawelew L, Malavasi S (2013) Sound production mechanism in Gobius paganellus (Gobiidae) The Journal of Experimental Biology 216: 3189-3199 Parmentier E, Lecchini D, Frederich B, Brié C, Mann D (2009) Sound production in four damselfish (Dascyllus) species: Phyletic relationships. Biological Journal of the Linnean Society 97: 928-940. Parmentier E, Petrinisec M, Fonseca PJ, Amorim MCP (2017) Sound-production mechanism in Pomatoschistus pictus. The Journal of Experimental Biology 220: 4374-4376. Pedroso SS, Barber I, Svensson O, Fonseca PJ, Amorim MCP (2013) Courtship sounds advertise species identity and male quality in sympatric Pomatoschistus spp gobies. PLoS One 8: e64620 Penzo E, Gandolfi G, Bargelloni L, Colombo L, Patarnello T (1998) Messinian salinity crisis and the origin of freshwater lifestyle in western Mediterranean Gobies. Molecular Biology and Evolution 15: 1472-1480. Percy D, Taylor G, Kennedy M (2006) Psyllid communication: acoustic diversity mate recognition and phylogenetic signal. Invertebrate Systematics 20: 431-445. Peters G, Tonkin-Leyhausen A (1999) Evolution of acoustic communication signals of mammals: friendly close-range vocalizations in Felidae (Carnivora). Journal of Mammal Evolution 6: 129-159. Popper AN (2011) Auditory System Morphology U: Farrell AP (ed.) Encyclopedia of fish physiology: from genome to environment vol 1. Academic Press San Diego 252-261 Pough F, Harvey CM, Heiser J (2009) Vertebrate life. 8th edn. Pearson. Preston JL (1978) Communication systems and social interactions in a goby-shrimp symbiosis. Animal Behaviour 26: 791-802. Přikryl T (2014) A new species of the sleeper goby (Gobioidei Eleotridae) from the České Středohoří Mountains (Czech Republic Oligocene) and analysis of the validity of the family Pirskeniidae. Paläontologische Zeitschrift 88: 187-196. Quignard JP, Tomasini JA (2000) Mediterranean fish biodiversity. Biologia Marina Mediterranea 7: 1-66

63 LITERATURA

Reid DF, Orlova MI (2002) Geological and evolutionary underpinnings for the success of Ponto-Caspian species invasions in the Baltic Sea and North American Great Lakes. Canadian Journal of Fisheries and Aquatic Sciences 59: 1144-1158. Renema W, Bellwood DR, Braga JC, Bromfield K, Hall R, Johnson KG, Lunt P, Meyer CP, McMonagle LB, Morley RJ, O’Dea A, Todd JA, Wesselingh FP, Wilson MEJ, Pandolfi JM (2008) Hopping hotspots: global shifts in marine biodiversity. Science 321: 654-657. Revell LJ, Harmon LJ, Collar DC (2008) Phylogenetic signal evolutionary process and rate. Systematic Biology 57: 591-601. Rice AN, Bass AH (2009) Novel vocal repertoire and paired swimbladders of the three-spined toadfish Batrachomoeus trispinosus: insights into the diversity of the Batrachoididae. The Journal of Experimental Biology 212: 1377-1391. Richardson WJ, Greene CR, Malme CJ, Thomson DH (1995) Marine Mammals and Noise San Diego. Academic Press Ritchie MG (2007) Sexual selection and speciation. Annual Review of Ecology, Evolution, and Systematics 38:79-102. Robillard T; Desutter-Grandcolas L (2004) High frequency calling in Eneopterinae crickets (Orthoptera Grylloidea Eneopteridae): adaptive radiation revealed by phylogenetic analysis. Biological Journal of the Linnean Society 83: 577- 584. Rogers PH, Cox M (1988) Underwater sound as a biological stimulus U: Fay RR, Popper AN, Tavolga WN, Atema J (ed.) Sensory biology of aquatic animals. Springer New York 131-149 Rögl F (1998) Paleogeographic consideration for Mediterranean and Paratethys seaways (Oligocene to Miocene). Annalen des Naturhistorischen Museums in Wien 99A: 279-310. Rollo A, Andraso G, Janssen J, Higgs D (2007) Attraction and localization of round goby (Neogobius melanostomus) to conspecific calls. Behaviour 144: 1-21. Šanda R, Kovačić M (2009) Freshwater gobies in the Adriatic drainage basin of the western Balkans. Annales, Series Historia Naturalis 19: 1-10. Sebastianutto L, Picciulin M, Costantini M, Rocca M, Ferrero E (2008) Four types of sounds from one winner: vocalizations during territorial behaviour in the red-mouthed goby Gobius cruentatus (Pisces Gobiidae). Acta Ethologica 11: 115-121. Seddon N, Merrill RM, Tobias JA (2008) Sexually selected traits predict patterns of species richness in a diverse clade of suboscine birds. The American Naturalist 171:620-631. Simmons A, Fay RR (2003) Acoustic Communication. Springer Simonović PD (1999) Phylogenetic relationships of Ponto- Caspian gobies and their relationship to the Atlantic-Mediterranean Gobiidae. Journal of Fish Biology 54: 533-555. Slabbekoorn H (2004) Singing in the wild: the ecology of birdsong. U: Marler P Slabbekoorn H (ed.) Nature’s Music: The Science of Birdsong. Elsevier Academic Press 178-205

64 LITERATURA

Smirnov AI (1986) Okuneobrazyne (bichkovidnye) skorpenoobraznye kambaloobraznye udilyishikoobrazye (Percifomes (Gobioidei) Scorpaeniformes Pleuronectiformes Lophiiformes). Fauna Ukraini 8: 7-183. Sorice M, Caputo V (1999) Genetic variation in seven goby species (Perciformes: Gobiidae) assessed by electrophoresis and taxonomic inference. Marine Biology 134: 327-333. Springer VG (1983) Tyson belos new genus and species of Western Pacific fish (Gobiidae Xenisthminae) with discussions of gobioid osteology and classification. Smithsonian Contributions to Zoology 390: 1-40. Stadler J M (2002) Evidence for a hydrodynamic mechanism of sound production by courting males of the notchtongue goby Bathygobius curacao (Metzlar). Bioacoustics 13: 145-152. Stepien CA, Tumeo MA (2006) Invasion genetics of Ponto-Caspian gobies in the Great Lakes: A “cryptic” species absence of founder effects and comparative risk analysis. Biological Invasions 8: 61-78. Sund O (1935) Echo sounding in fishery research. Nature 135: 953 Suthers RA, Fitch WT, Fay RR, Popper AN (2016) Vertebrate Sound Production and Acoustic Communication. Springer International Publishing Takemura A (1984) Acoustical behaviour of the freshwater goby Odontobutis obscura. Bulletin of the Japanese Society for the Science of Fish 50: 561-564. Tavolga WM (1956) Visual chemical and sound stimuli as cues in the sex discriminatory behaviour of the gobiid fi sh Bathygobius soporator. Zoologica 41: 49-65 Thacker CE (2000) Phylogeny of the Wormfishes (Teleostei: Gobioidei: Microdesmidae). Copeia 2000: 940-957. Thacker CE (2009) Phylogeny of Gobioidei and placement within Acantomorpha with a new classification and investigation of diversification and character evolution. Copeia 2009: 93- 104. Thacker CE (2014) Species and shape diversification are inversely correlated among gobies and cardinalfishes (Teleostei: Gobiiformes). Organisms Diversity and Evolution 14: 419-436. Thacker CE (2015) Biogeography of goby lineages (Gobiiformes: Gobioidei): Origin invasions and extinction throughout the Cenozoic. Journal of Biogeography 42: 1615-1625. Thacker CE, Roje DM (2011) Phylogeny of Gobiidae and identification of gobiid lineages. Systematics and Biodiversity 9: 329-347. Thinh VN, Hallam C, Roos C, Hammerschmidt K (2011) Concordance between vocal and genetic diversity in crested gibbons. BMC Evolutionary Biology 11: 36. Thompson KS (1968) Lung ventilation in dipnoan fishes. Postila 122: 1-6. Toews D, Irwin D (2008) Cryptic speciation in a Holarctic passerine revealed by genetic and bioacoustic analyses. Molecular Ecology 17: 2691-705.

65 LITERATURA

Tornabene L, Chen Y, Pezold F (2013) Gobies are deeply divided: phylogenetic evidence from nuclear DNA (Teleostei: Gobioidei: Gobiidae). Systematics and Biodiversity 11:1-17. Torricelli P, Lugli M, Gandolfi G (1985) A quantitative analysis of the fanning activity in the male Padogobius martensi (Pisces: Gobiidae). Behaviour 92: 288-301 Torricelli P, Lugli M, Gandolfi G (1986) A quantitative analysis of the occurrence of visual and acoustic displays during the courtship in the freshwater goby Padogobius martensi (Günther 1961) (Pisces Gobiidae). Bollettino di zoologia 53: 85-89. Torricelli P, Lugli M, Pavan G (1990) Analysis of sounds produced by male Padogobius martensi (pisces Gobiidae) and factors affecting their structural properties. Bioacoustics 2: 261-275. Urick RJ (1983) Principles of Underwater Sound. McGraw-Hill New York Vandewalle P, Seiller P, Chardon M (1982) Particularités anatomiques et fonctionnelles de la région céphalique de Blennius pholis L (Pisces Blenniidae). Cybium 6: 73-94. Velásquez JN, Marambio J, Brunetti E, Méndez MA, Vásquez RA, Penna M (2013) Bioacoustic and genetic divergence in a frog with a wide geographical distribution. Biological Journal of the Linnean Society 110: 142-155 Wiley RH, Richards DG (1982) Adaptations of acoustic communication in birds: sound transmission and signal detection. Acoustic communication in birds. Academic Press New York 131-278 Wilkins M, Seddon N, Safran R (2012) Evolutionary divergence in acoustic signals: Causes and consequences Trends in Ecology and Evolution 28: 156-166. Winterbottom R (1974) A descriptive synonymy of the striated muscles of the Teleostei. Proceedings of the Academy of Natural Sciences of Philadelphia 125: 225-317. Winterbottom R, Alofs KM, Marseu A (2011) Life span growth and mortality in the western Pacific goby Trimma benjamini and comparisons with T. nasa. Environmental Biology of Fishes 91: 295-301. Wood R (1999) Reef evolution. Oxford University Press Oxford UK. Wysocki LE, Ladich F (2001) The ontogenetic development of auditory sensitivity vocalization and acoustic communication in the labyrinth fish Trichopsis vittata. Journal of Comparative Physiology A 187: 177-187 Wysocki LE, Ladich F (2002) Ontogeny of hearing and sound production in fishes. Bioacoustics 12: 183-185. Zeyl JN Malavasi S Holt DE Noel P Lugli M i Johnston CE (2016) Convergent aspects of acoustic communication in darters sculpins and gobies. U: Sisneros AJ (ed.) Fish Hearing And Bioacoustics: An Anthology In Honor Of Arthur N Popper And Richard R Fay. Springer https://www.britannica.com/animal/fish (zadnji put pristupljeno 28. kolovoza 2020.) http://www-9.unipv.it/cibra/edu_underwaterbioacoustics_uk.html (zadnji put pristupljeno 28. kolovoza 2020.)

66 ŽIVOTOPIS

6. ŽIVOTOPIS

67 ŽIVOTOPIS ŽIVOTOPIS

Sven Horvatić rođen je 8. svibnja 1991. godine u Zagrebu. Nakon završene osnovne škole u Rugvici, 2006. godine upisuje se u Opću gimnaziju u Dugom Selu koju završava 2010. godine. Iste godine upisuje se na Prirodoslovno-matematički fakultet (PMF) u Zagrebu, na preddiplomski smjer Znanosti o okolišu. Titulu sveučilišnog prvostupnika struke znanosti o okolišu stječe 2013. godine te se iste godine upisuje na diplomski smjer Eksperimentalne biologije (modul Zoologija) na PMF-u u Zagrebu. Diplomski rad „Produkcija zvukova pontokaspijskog riječnog glavočića (Neogobius fluviatilis Pallas, 1814) i njegova srodnost s mediteranskim vrstama s obzirom na akustična svojstva“ izradio je 2015. godine te ga uspješno obranio pod mentorstvom prof. dr. sc. Davora Zanelle. Tijekom diplomskog studija sudjelovao je u programu ERASMUS studentske mobilnosti, u sklopu kojeg je proveo stručnu praksu u ljetnom semestru 2014. godine na Sveučilištu Ca' Foscari di Venezia. Za vrijeme stručne prakse radio je pod mentorstvom prof. dr. sc. Stefana Malavasija, tijekom koje je većina znanstvene pozornosti bila posvećena provedbi laboratorijskih bioakustičkih eksperimenata i akustičkoj obradi zvukova europskih glavoča porodice Gobiidae. Poslije završetka diplomskog studija 2015. godine, upisuje prvu godinu poslijediplomskog studija Biologije na PMF-u u Zagrebu pod mentorstvom prof. dr. sc. Davora Zanelle. Godinu dana poslije zaposlio se na Biološkom odsjeku PMF-a u Zagrebu kao znanstveni novak - asistent na stručnom projektu „Provođenje programa praćenja stanja u slatkovodnom ribarstvu u 2016. godini (grupa Es – Ribolovno područje Jadran)“. Od 2017. godine radi kao stručni suradnik na projektu „Biološka ispitivanja nadzemnih voda na HE Varaždin, HE Čakovec i HE Dubrava“ na Zoologijskom zavodu Biološkog odsjeka PMF-a. U sklopu poslijediplomskog studija sudjelovao je na nekoliko znanstvenih usavršavanja iz područja bioakustike izvan Hrvatske: u Italiji (Ca' Foscari Sveučilište u Veneciji, mentor: prof. Stefano Malavasi) i Belgiji (Sveučilište u Liègeu, mentor: prof. Eric Parmentier). Autor je 13 izvornih znanstvenih radova (10 u CC bazi po WOS-u), dok je na njih šest prvi autor. Također, usmeno je prezentirao dva izlaganja na međunarodnim kongresima (Švicarska i Njemačka), dok je na jednom koautor (Zagreb). Prvi je autor na tri posterska izlaganja prezentirana na međunarodnim kongresima (Švicarska, Bosna i Hercegovina i Hrvatska). Od 2016. godine vodi praktikumsku nastavu iz kolegija Vertebrata, Kralješnjaci, Zoologija 3, Terenska nastava i Ihtiologija i ribarstvo slatkih voda, za studente preddiplomskih, diplomskih i integriranih preddiplomskih i diplomskih studija na PMF-u.

68 ŽIVOTOPIS

ZNANSTVENE PUBLIKACIJE: Horvatić, S., Bem, L., Malavasi, S., Marčić, Z., Buj, I., Mustafić, P., Ćaleta, M., Zanella, D. Comparative Tutman, P., Zanella, D., Horvatić, S., Hamzić, A., analysis of sound production between the bighead goby Adrović, A., Dulčić, J., Glamuzina, B. Freshwater gobies Ponticola kessleri and the round goby Neogobius (Gobiidae) of Bosnia and Herzegovina: a review of the melanostomus: implications for phylogeny and current status and distribution. Journal of Vertebrate systematics. Environmental Biology of Fishes. 102 Biology. 69 (2020), 4; 1-15. (2019), 5; 1-13. Piria, M., Zanella, D., Marčić, Z., Ćaleta, M., Horvatić, Horvatić, S., Malavasi, S., Parmentier, E., Marčić, Z., S., Jakšić, G., Buj, I., Paunović, M., Simonović, P., Buj, I., Mustafić, P., Ćaleta, M., Smederevac-Lalić, M., Mustafić, P. Has the racer goby Babka gymnotrachelus Skorić, S. i Zanella, D. Acoustic communication during (Kessler, 1857) failed to invade the Danube tributaries, reproduction in the basal gobioid Amur sleeper and the the Sava and Drava Rivers? Fundamental and Applied putative sound production mechanism. Journal of Limnology. (2020) In press Zoology. 309 (2019), 4; 269-279. Buj, I., Raguž, L., Marčić, Z., Ćaleta, M., Duplić, A., Zanella, D.; Marčić, Z.; Ćaleta, M.; Buj, I.; Zrnčić, S.; Zanella, D., Mustafić, P., Ivić, L., Horvatić, S., Karlović Horvatić, S.; Mustafić, P. Early development of the R. Plitvice Lakes National park harbors ancient, yet freshwater goby Orsinogobius croaticus endemic to endangered diversity of trout (genus Salmo). Journal of Croatia and Bosnia-Herzegovina. Cybium. 41 (2017), 4; Applied Ichthyology. (2020) In press 335-342. Marčić, Z., Abramović, A., Ćaleta, M., Buj, I., Zanella, Horvatić, S., Marčić, Z., Mrakovčić, M., Mustafić, P., D., Horvatić, S., Mustafić, P. Early development of the Buj, I., Ćaleta, M., Zanella, D. Threatened Fishes of the endemic dace Telestes karsticus (Leuciscidae, World: Orsinogobius croaticus (Mrakovčić, Kerovec, Cypriniformes) in a Dinaric karst stream in Croatia. Mišetić and Schneider, 1996) (Teleostei: Gobiidae). Journal of applied ichthyology. (2020) In press Ribarstvo. 75 (2017), 1; 23-30. Buj, I., Marčić, Z., Čavlović, K., Ćaleta, M., Tutman, P., Mustafić, P., Buj, I., Opašić, M., Zanella, D., Marčić, Z., Zanella, D., Duplić, Al., Raguž, L., Ivić, L., Horvatić, S., Ćaleta, M., Šanda, R., Horvatić, S., Mrakovčić, M. Mustafić, P. Multilocus phylogenetic analysis helps to Morphological comparison of Delminichthys ghetaldii untangle the taxonomic puzzle of chubs (genus Squalius: (Steindachner, 1882), D. adspersus (Heckel, 1843), D. Cypriniformes: Actinopteri) in the Adriatic basin of jadovensis (Zupančič & Bogutskaya, 2002) and D. Croatia and Bosnia and Herzegovina. Zoological Journal krbavensis (Zupančič & Bogutskaya, 2002), endemic of the Linnaean Society. 189 (2020) 3; 953-974. species of the Dinaric karst, Croatia. Journal of Applied Buj, I, Mustafić, P., Ćaleta, M., Marčić, Z., Ivić, L., Ichthyology. 33 (2017), 2; 256-262. Žalac, S., Zanella, D., Karlović, R., Horvatić, S., Raguž, Horvatić, S., Cavraro, F., Zanella, D., Malavas, S. Sound L. Peculiar occurrence of Cobitis bilineata Canestrini, production in the Ponto-Caspian goby Neogobius 1865 and Sabanejewia larvata (De Filippi, 1859) fluviatilis and acoustic affinities within the Gobius (Cobitidae, Actinopteri) in the Danube River basin in lineage: implications for phylogeny. Biological Journal of Croatia. Fundamental and Applied Limnology. (2020) In the Linnaean Society. 117 (2016), 3; 564-573. press SAŽETCI S KONFERENCIJA: Ćaleta, M., Marčić, Z., Buj, I., Zanella, D., Mustafić, P., Duplić, A., Horvatić, S. A review of extant croatian Horvatić, S., Špelić, A., Amorim, MCP,. Fonseca, PJ., freshwater fish and lampreys, Annotated list and Zanella, D. Endemic goby calling: acoustics and vocal distribution. Croatian Journal of Fisheries. (2019) 77; behavior of Orsinigobius croaticus. Goby meeting 2020. 137-234. (2020)

69 ŽIVOTOPIS

Piria, M., Zanella, D., Marčić, Z., Ćaleta, M., Horvatić, Mustafić i sur., Prirodoslovno-matematički fakultet, S., Jakšić, G., Buj, I., Paunović, M., Simonović, P., Biološki odsjek, Zagreb. Mustafić, P. Has the racer goby Babka gymnotrachelus Research and optimisation of ichthyocenosis of Butoniga (Kessler, 1857) failed to invade the Danube tributaries, Reservoir for the purpose of reducing the trophic level in the Sava and Drava Rivers? 11th Symposium for 2017, 2018 and 2019. Perica Mustafić i sur., European Freshwater Sciences. (2019) Prirodoslovno-matematički fakultet, Biološki odsjek, Horvatić, S., Zanella. D., Malavasi, S. Evolution of Zagreb. acoustic communication in European gobies (Gobius Usluga istraživanja riba i školjkaša u rijeci Sutli u lineage) based on molecular markers. XVI European projektu Veze prirode (Fishes and Unionidae of Sutla Congress of Ichthyology. (2019) River). Zoran Marčić i sur., Prirodoslovno-matematički Horvatić, S., Malavasi, S., Parmentier, E., Marčić, Z., fakultet, Biološki odsjek, Zagreb. Buj, I., Mustafić, P., Ćaleta, M., Smederevac-Lalić, M., Stručna podloga za Plan upravljanja (s akcijskim planom) Skorić, S. i Zanella, D. Acoustics, vocal behavior and za vrste roda Salmo (982617). Ivana Buj i sur., Hrvatsko putative sound producing mechanism in Amur sleeper ihtiološko društvo, Zagreb. Perccottus glenii. XVI European Congress of Ichthyology. (2019) Usluga izrade stručne podloge – svijetlica i kapelska svijetlica u sklopu projekta „Izrada prijedloga Planova Horvatić, S., Bem, L., Malavasi, S., Marčić, Z., Buj, I., upravljanja strogo zaštićenim vrstama (s akcijskim Mustafić, P., Ćaleta, M., Zanella, D. Acoustic repertoire planovima)“. (178923). Zoran Marčić i sur., Hrvatsko of bighead goby Ponticola kessleri (Günther 1861) ihtiološko društvo, Zagreb. (Gobiidae), and implications on sound production within Ponto-Caspian species. 1st SouthEast European RADIONICE, ZNANSTVENO USAVRŠAVANJE I SURADNJA: Ichthyological Conference. (2017) Training workshop on fish. The International Horvatić, S., Marčić, Z., Mrakovčić, M., Mustafić, P., Commission for the Protection of the Danube River Buj, I., Ćaleta, M., Zanella, D. Ichthyofauna structure of (ICPDR). (2018) the Save River within the riprap zone as a specific habitat type, with emphasis on the presence of the round goby Znanstveno usavršavanje izvan Hrvatske, prekogranična Neogobius melanostomus (Pallas, 1814). FINS II akademska mobilnosti prema visokoškolskim i conference. (2016) znanstvenim ustanovama, Sveučilište u Liègeu, Belgija. (2019) ZNANSTVENI PROJEKTI Znanstveno usavršavanje izvan Hrvatske, Sveučilište Ca' Utjecaj klimatskih promjena na bioraznolikost koralja - Foscari u Veneciji, Italija (2019, 2018, 2017) istraživanje slučaja masovnih ugibanja u Jadranskom moru“ (IP-2019-04-3389). Nositelj: Izv. prof. dr. sc. Petar Upoznavanje sa sintaksom jezika R i njegova primjena u Kružić, Prirodoslovno-matematički fakultet, Biološki osnovnoj statističkoj i grafičkoj analizi podataka (S720). odsjek, Zagreb. Hrvatska zaklada za znanost (HRZZ). Sveučilište u Zagrebu , Sveučilišni računski centar. (2018) STRUČNI PROJEKTI NASTAVNA AKTIVNOSTI I INSTITUCIJSKA Biological study of surface water of the hydropower ZADUŽENJA: systems Varaždin, Čakovec and Dubrava in 2016, 2017, Prirodoslovno-matematički fakultet, Sveučilište u 2018 and 2019. Perica Mustafić i sur., Prirodoslovno- Zagrebu. Biološki odsjek. Kolegij: Kralješnjaci (40873) - matematički fakultet, Biološki odsjek, Zagreb. asistent (2016 – danas) Monitoring project of Croatian freshwater ichthyofauna in 2016, 2017, 2018 and 2019 (zone E – Adriatic). Perica

70 ŽIVOTOPIS

Prirodoslovno-matematički fakultet, Sveučilište u Zagrebu. Biološki odsjek. Kolegij: Vertebrata (40923) - asistent (2016 – danas)

Prirodoslovno-matematički fakultet, Sveučilište u Zagrebu. Biološki odsjek. Kolegij: Zoologija 3 (40896) - asistent (2016 – danas)

Prirodoslovno-matematički fakultet, Sveučilište u Zagrebu. Biološki odsjek. Kolegij: Terenska nastava (40876) - asistent (2016 – danas)

Prirodoslovno-matematički fakultet, Sveučilište u Zagrebu. Biološki odsjek. Kolegij: Terenska nastava iz biološke, geografske i geološke zaštite okoliša (73836) - asistent (2016 – danas)

Prirodoslovno-matematički fakultet, Sveučilište u Zagrebu. Biološki odsjek. Kolegij: Ihtiologija i ribarstvo slatkih voda (60228) - asistent (2019 – danas)

- predstavnik asistenata i poslijedoktoranata na Vijeću Biološkog odsjeka Prirodoslovno-matematičkog fakulteta (2018 – danas)

- zamjenik predstavnika asistenata i poslijedoktoranata na Vijeću Kolegija Biološkog odsjeka (2018 – danas)

- član Studentskog zbora Prirodoslovno-matematičkog fakulteta (2019 – danas)

- član organizacijskog odbora manifestacije “Noć Biologije (Dan i noć PMF-a 2019)”. (2019)

- član Hrvatskog ihtiološkog društva (HID) (2016 – danas)